CN106156880A - A kind of predict the method for inventory allocation ratio, device and electronic equipment - Google Patents
A kind of predict the method for inventory allocation ratio, device and electronic equipment Download PDFInfo
- Publication number
- CN106156880A CN106156880A CN201510194457.5A CN201510194457A CN106156880A CN 106156880 A CN106156880 A CN 106156880A CN 201510194457 A CN201510194457 A CN 201510194457A CN 106156880 A CN106156880 A CN 106156880A
- Authority
- CN
- China
- Prior art keywords
- ratio
- commodity
- prediction
- warehouse
- default
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
This application discloses and a kind of predict the method for inventory allocation ratio, device and electronic equipment.The method of wherein said prediction inventory allocation ratio includes: obtain the commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated, as prediction characteristic;According to described prediction characteristic and the forecast model previously generated, generate the described commodity to be allocated predictive value in the inventory allocation ratio in described warehouse to be allocated.Use the method that the application provides, the combined factors such as historical sales behaviors based on commodity and issuable purchasing demand in the future can assess the commodity to be allocated allocation proportion in warehouse to be allocated, therefore, improve the prediction accuracy of inventory allocation ratio, avoid too much relying on manpower and carry out inventory allocation, thus reduce Out of Stock rate and dispensing timeliness, improve stock's turnover level and the effect of consumer experience.
Description
Technical field
The application relates to data mining technology field, be specifically related to a kind of method predicting inventory allocation ratio and
Device.The application relates to a kind of electronic equipment simultaneously.
Background technology
Along with the development of Internet technology, occur in that one carries out business by public computer communications network
The modern business mode of business activity, it may be assumed that ecommerce.Ecommerce breach conventional business the time,
Restriction on territory, has safe and reliable, quick, clear and definite and convenient feature, has been increasingly becoming society
The major way of life.Ecommerce can be carried out whenever and wherever possible by international interconnection network (INTERNET) to live
Dynamic.
B2C (Business to Consumer, business to consumer) pattern is a kind of typical ecommerce mould
Formula.B2C e-commerce website is the most.It is traded by B2C pattern, relates generally on three lines
Role: platform operator, stand interior operator, consumer.Wherein, platform operator be responsible for ecommerce put down
The operation of platform (hereinafter referred to as transaction platform);In standing, operator is also known as enterprise or businessman, is responsible for providing commodity;
Consumer orders, by transaction platform, the commodity that businessman provides.Except role on line, B2C pattern further relates to line
Lower role, such as: logistics distribution.B2C pattern is substantially on line under order, line dispensing.
Identical with conventional business, the commodity that each businessman of ecommerce is managed also are to leave all parts of the country in
Storehouse in.Businessman in conventional business usually in accordance with it in the recent period in the sales data in each warehouse, for respectively
Individual storehouse is equipped with appropriate number of commodity.At present, each businessman of ecommerce is as conventional business, also
It is in the recent period in the sales data in each warehouse according to it, is equipped with appropriate number of commodity for each storehouse.But,
This recent sales data according only to each warehouse, carry out the method for inventory allocation have a disadvantage in that by
Excessively simple in the historic sales data of its foundation, the accuracy causing inventory allocation is low, thus brings business
Family's problem that supply chain efficiency is low and consumer experience is poor.
Therefore, prior art existence carries out inventory allocation according to commodity in the recent sales data of each storehouse
The problem that quantity allotted accuracy is low.
Summary of the invention
The application provides a kind of method and apparatus predicting inventory allocation ratio, there is root solving prior art
Carry out, in the recent sales data of each storehouse, the problem that the quantity allotted accuracy of inventory allocation is low according to commodity.
The application additionally provides a kind of electronic equipment.
The application provides a kind of method predicting inventory allocation ratio, including:
Obtain the commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated, as prediction feature
Data;
According to described prediction characteristic and the forecast model previously generated, generate described commodity to be allocated and exist
The predictive value of the inventory allocation ratio in described warehouse to be allocated.
Optionally, described forecast model uses following steps to generate:
Obtain each commodity the history feature data relevant to sales volume in each warehouse and with described history feature
The inventory allocation ratio that data are corresponding--i.e. effective sale ratio, as training set;
By machine learning algorithm, obtain described forecast model from described training set learning.
Optionally, described machine learning algorithm includes regression algorithm.
Optionally, described regression algorithm includes linear regression algorithm, regression tree algorithm, iteration decision tree
Algorithm or random forests algorithm.
Optionally, described obtain described forecast model by machine learning algorithm from described training set learning,
Including:
The training initial predicted of inventory allocation ratio it is respectively provided with for every training data in described training set
Value;
According to described training initial prediction and described training set, use iteration decision Tree algorithms, generate institute
State forecast model.
Optionally, the effective sale of the preset times before described training initial prediction uses current number of times
The meansigma methods of ratio.
Optionally, described according to described prediction characteristic with the forecast model that previously generates, generate described
Commodity to be allocated at the predictive value of the inventory allocation ratio in described warehouse to be allocated, including:
The prediction initial predicted of inventory allocation ratio is set in described warehouse to be allocated for described commodity to be allocated
Value;
According to described prediction characteristic, prediction initial prediction and the forecast model previously generated, raw
Become described commodity to be allocated at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
Optionally, the effective sale of the preset times before described prediction initial prediction uses current number of times
The meansigma methods of ratio.
Optionally, described obtain described forecast model by machine learning algorithm from described training set learning,
In the following way:
According to default various machine learning algorithms, obtain default each with described from described training set learning
Plant the forecast model that machine learning algorithm is the most corresponding.
Optionally, described according to described prediction characteristic with the forecast model that previously generates, generate described
Commodity to be allocated at the predictive value of the inventory allocation ratio in described warehouse to be allocated, including:
To described prediction characteristic, each forecast model being respectively adopted study acquisition carries out described stock
The prediction of allocation proportion, it is thus achieved that the predictive value of the described inventory allocation ratio that each forecast model dopes;
The weight preset respectively according to the predictive value doped for each forecast model, to each prediction mould described
The predictive value of the described inventory allocation ratio that type dopes is weighted average computation, as described business to be allocated
Product are at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
Optionally, described acquisition commodity to be allocated are at the characteristic relevant to sales volume in warehouse to be allocated, bag
Include:
Obtain original relevant to sales volume in described warehouse to be allocated of described commodity to be allocated in preset time range
Data;
According to the described initial data relevant to sales volume and the computing formula of described characteristic, calculate and obtain institute
State the characteristic relevant to sales volume.
Optionally, the described characteristic relevant to sales volume includes at least one of following data:
Preset the sale in preset time range of the default commodity in warehouse and click on ratio;Described sale is clicked on
Ratio, refers to that the sale number of clicks presetting the default commodity in warehouse sells click time the total of default commodity
Ratio in number;
Preset the sale in preset time range of the default commodity in warehouse and click on buyer's ratio;Described sale
Click on buyer's ratio, refer to that the selling of default commodity presetting in warehouse clicks on buyer's number at default commodity
Total ratio sold in click buyer's number;
Preset the click of taking in preset time range of the default businessman belonging to default commodity in warehouse to compare
Example;Described take click ratio, refer to preset taking a little of the default businessman belonging to default commodity in warehouse
Hit number of times, in the ratio always taken in number of clicks of the default businessman belonging to default commodity;
Preset the taking in preset time range of the default businessman belonging to default commodity in warehouse and click on buyer
Ratio;Described take click buyer's ratio, refer to preset the default businessman belonging to default commodity in warehouse
Take click buyer's number, at the ratio always taken in click buyer's number of the default businessman belonging to default commodity
Example;
Preset the sale ratio of the merchandise classification belonging to default commodity presetting businessman belonging to the default commodity in warehouse
Example, sells ratio as second;Described second sells ratio, refers to preset belonging to the default commodity in warehouse
Preset the sales volume of the merchandise classification belonging to default commodity of businessman, belonging to default commodity, preset businessman
Preset the ratio in total sales volume of the merchandise classification belonging to commodity;
The sale ratio of the merchandise classification belonging to default commodity in default warehouse, sells ratio as the 3rd;
Described 3rd sells ratio, refers to preset the sales volume of the merchandise classification belonging to default commodity in warehouse,
Ratio in total sales volume of the merchandise classification belonging to default commodity;
Preset the commodity total sales volume in preset time range;
Preset the shopping cart in preset time range of the default commodity in warehouse and click on ratio;Described shopping cart
Click ratio, refers to the shopping cart number of clicks presetting the default commodity in warehouse total shopping at default commodity
Ratio in car number of clicks;
Preset the shopping cart in preset time range of the default commodity in warehouse and click on buyer's ratio;Described purchase
Buyer's ratio clicked on by thing car, refers to that the shopping cart presetting the default commodity in warehouse clicks on buyer's number default
The ratio in buyer's number clicked on by total shopping cart of commodity;
Preset the collection in preset time range of the default businessman belonging to default commodity in warehouse and click on ratio
Example;Described collection clicks on ratio, refers to preset the collection number of clicks of the default businessman belonging to commodity,
Ratio in total collection number of clicks of the default businessman belonging to default commodity;
Preset the collection click in preset time range of the default businessman belonging to default commodity in warehouse to buy
Family's ratio;Described collection clicks on buyer's ratio, refers to preset the collection point of the default businessman belonging to commodity
Hitting buyer's number, the total collection in the default businessman belonging to default commodity clicks on the ratio in buyer's number;
The predetermined ratio of the default commodity in default warehouse;Described predetermined ratio, refer to preset in warehouse is pre-
If the ratio that the predetermined quantity of commodity is in total predetermined quantity of default commodity.
Optionally, the described characteristic relevant to sales volume also includes:
The sale ratio of the default commodity in default warehouse, sells ratio as first;Described first sells ratio
Example, refers to the sales volume presetting the default commodity in warehouse ratio in total sales volume of default commodity.
Accordingly, a kind of device predicting inventory allocation ratio of the application, including:
Acquiring unit, for obtaining the commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated,
As prediction characteristic;
Predicting unit, for according to described prediction characteristic and the forecast model previously generated, generating institute
State the commodity to be allocated predictive value in the inventory allocation ratio in described warehouse to be allocated.
Optionally, also include:
Training unit, is used for generating described forecast model.
Optionally, described training unit includes:
Obtain subelement, for obtaining each commodity history feature data relevant to sales volume in each warehouse
And the inventory allocation ratio corresponding with described history feature data--i.e. effective sale ratio, as training set;
Study subelement, for by machine learning algorithm, obtains described prediction from described training set learning
Model.
Optionally, described study subelement includes:
Initial value arranges subelement, divides for being respectively provided with stock for every training data in described training set
The training initial prediction of proportioning example;
Generate subelement, for according to described training initial prediction and described training set, use iteration certainly
Plan tree algorithm, generates described forecast model.
Optionally, the effective sale of the preset times before described training initial prediction uses current number of times
The meansigma methods of ratio.
Optionally, described predicting unit includes:
Initial value arranges subelement, divides for arranging stock for described commodity to be allocated in described warehouse to be allocated
The prediction initial prediction of proportioning example;
Computation subunit, for according to described prediction characteristic, prediction initial prediction and pre-Mr.
The forecast model become, generates the prediction in the inventory allocation ratio in described warehouse to be allocated of the described commodity to be allocated
Value.
Optionally, described obtain described forecast model by machine learning algorithm from described training set learning,
In the following way:
According to default various machine learning algorithms, obtain default each with described from described training set learning
Plant the forecast model that machine learning algorithm is the most corresponding.
Optionally, described predicting unit includes:
Prediction subelement, for described prediction characteristic, being respectively adopted each prediction that study obtains
Model carries out the prediction of described inventory allocation ratio, it is thus achieved that the described inventory allocation that each forecast model dopes
The predictive value of ratio;
Computation subunit, for the weight preset respectively according to the predictive value doped for each forecast model,
The predictive value of the described inventory allocation ratio doping each forecast model described is weighted average computation,
As described commodity to be allocated at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
Optionally, described acquiring unit includes:
Obtaining subelement, in being used for obtaining preset time range, described commodity to be allocated are in described warehouse to be allocated
The initial data relevant to sales volume;
Computation subunit, for according to the described initial data relevant to sales volume and the calculating of described characteristic
Formula, calculates and obtains the described characteristic relevant to sales volume.
Additionally, the application also provides for a kind of electronic equipment, including:
Display;
Processor;And
Memorizer, for the device of Storage Estimation inventory allocation ratio, the dress of described prediction inventory allocation ratio
Put when being performed by described processor, comprise the steps: to obtain commodity to be allocated warehouse to be allocated with pin
The characteristic that amount is relevant, as prediction characteristic;According to described prediction characteristic and pre-Mr.
The forecast model become, generates the prediction in the inventory allocation ratio in described warehouse to be allocated of the described commodity to be allocated
Value.
Compared with prior art, the application has the advantage that
The prediction method of inventory allocation ratio that the application provides, device and electronic equipment, treat point by obtaining
Join the commodity characteristic relevant to sales volume in warehouse to be allocated, and according to the characteristic obtained and in advance
The forecast model generated, generates the commodity to be allocated predictive value in the inventory allocation ratio in warehouse to be allocated.By
The method provided in the application is historical sales behavior based on commodity and issuable purchasing demand in the future
Assess the commodity to be allocated allocation proportion in warehouse to be allocated etc. combined factors, this improves inventory allocation
The prediction accuracy of ratio, it is to avoid too much rely on manpower and carry out inventory allocation, thus reduce Out of Stock
Rate and dispensing timeliness, improve stock's turnover level and the effect of consumer experience.
Accompanying drawing explanation
Fig. 1 is the flow chart of the embodiment of the method for the prediction inventory allocation ratio of the application;
Fig. 2 is the idiographic flow of step S101 of the embodiment of the method for the prediction inventory allocation ratio of the application
Figure;
Fig. 3 is the flow chart generating forecast model of the embodiment of the method for the prediction inventory allocation ratio of the application;
Fig. 4 is the idiographic flow of step S303 of the embodiment of the method for the prediction inventory allocation ratio of the application
Figure;
Fig. 5 is the schematic diagram of the forecast model of the embodiment of the method generation of the prediction inventory allocation ratio of the application;
Fig. 6 is the schematic diagram of the embodiment of the method prediction process of the prediction inventory allocation ratio of the application;
Fig. 7 is the schematic diagram of the device embodiment of the prediction inventory allocation ratio of the application;
Fig. 8 is the concrete schematic diagram of the device embodiment of the prediction inventory allocation ratio of the application;
Fig. 9 is the schematic diagram of the electronic equipment embodiment of the application.
Detailed description of the invention
Elaborate a lot of detail in the following description so that fully understanding the application.But the application
Can implement to be much different from alternate manner described here, those skilled in the art can without prejudice to
Doing similar popularization in the case of the application intension, therefore the application is not limited by following public being embodied as.
In this application, it is provided that a kind of method and apparatus predicting inventory allocation ratio.In following enforcement
Example is described in detail one by one.
The method of the prediction inventory allocation ratio that the application provides, its basic thought is: according to commodity in warehouse
In the characteristic relevant with sales volume, use training in advance generate forecast model, it was predicted that commodity are in warehouse
The inventory allocation ratio of middle following a period of time.Due to the application provide method be according to commodity in warehouse
Various dimensions characteristic set out, it was predicted that commodity inventory allocation ratio in warehouse, therefore, it is possible to improve storehouse
Deposit the prediction accuracy of allocation proportion.
Refer to Fig. 1, it is the flow chart of embodiment of the method for prediction inventory allocation ratio of the application.Described
Method comprises the steps:
Step S101: obtain the commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated, as
Prediction characteristic.
The method of the prediction inventory allocation ratio that the application provides reassigns ratio to stock in future and estimates, both
Do not get sth into one's head, be not as prior art inventory allocation ratio according only to the past directly give
Stock reassigns ratio, and on the basis of being built upon issuable purchasing demand in the future.Due in reality
In application, demand for commodity has certain successional feature, and inventor finds, can from commodity transaction daily record with
And the data source such as inventory allocation data processes the data that sales volume produces impact, it may be assumed that relevant to sales volume
Characteristic, according to these characteristics and the forecast model previously generated, it is possible to increase commodity stocks is distributed
The accuracy of scale prediction value.
The characteristic relevant to sales volume described herein, including the predictive value possibility to inventory allocation ratio
Produce the various characteristics of impact.The different characteristics power of influence to the predictive value of inventory allocation ratio
Differing, some of which characteristic produces main impact to the predictive value of inventory allocation ratio,
And other characteristics produce more side effect to the predictive value of inventory allocation ratio.It is set forth below out
Characteristic conventional in the present embodiment.
1) characteristic relevant to daily sales volume
The characteristic relevant to daily sales volume described in the embodiment of the present application, including each commodity in each storehouse
The history in storehouse sells IPV/UV/ number of packages accounting in 7 days, 14 days, 30 days, 60 days;Each supplier is each
The history in warehouse takes IPV/UV/ number of packages accounting in 7 days, 14 days, 30 days, 60 days;Each supplier two grades
The classification history in each warehouse sells number of packages accounting in 7 days, 14 days, 30 days, 60 days;The basic letter of commodity
Breath, such as two grades of classifications, leaf classification etc.;And commodity are at history 7 days, 14 days, 30 days, 60 days total pins
Amount etc..
The quantity of the UV (Unique Visitor) described in the embodiment of the present application refers to independent access person.Access one
One computer client of individual on-line shop is a visitor, it may be assumed that an independent access person, same of 0:00-24:00
Computer client only can be recorded once.Described IPV (Page Views) is that browsing of item detail page is secondary
Number, it may be assumed that after buyer finds the dotey in shop, clicks through the number of times of dotey's details page, namely single product is clear
The amount of looking at.Described IPV_UV is the independent access person of browsed commodity details.
In actual applications, can be tied by Merchant sales behavior and customer buying behavior and network topology
Structure, obtains initial data characteristic involved by relevant to daily sales volume, and such as, commodity are in each storehouse
IPV, UV data taking and selling in storehouse, commodity historical sales data, two grades of classifications of supplier are sold
Data and the essential information etc. of commodity.
2) characteristic relevant to following sales volume
The characteristic relevant to following sales volume described in the embodiment of the present application, including each commodity in each storehouse
The history in storehouse 7 days, 14 days, 30 days, 60 days shopping cart IPV/UV/ accountings;Each commodity are in each warehouse
History 7 days, 14 days, 30 days, 60 days collection IPV/UV/ accountings;Presell commodity presell deposit is often
Individual warehouse accounting etc..
The application to be implemented provide method, it is necessary first to obtain commodity to be allocated warehouse to be allocated with pin
The characteristic that amount is relevant, as the various dimensions characteristic of prediction.Described in the embodiment of the present application
Commodity to be allocated, in the characteristic relevant to sales volume in warehouse to be allocated, can be i.e. to previously generate and store
Good characteristic, it is also possible to be according to the initial data relevant to sales volume and characteristic computing formula, real
Time calculate generate characteristic.
In actual applications, if previously generated described commodity to be allocated warehouse to be allocated and sales volume
Relevant characteristic, the most only need to be logical with described commodity to be allocated and described warehouse to be allocated as search condition
Cross data retrieval method, obtain the described commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated.
Without previously generating the described commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated, then
Need according to the initial data relevant to the sales volume of described commodity to be allocated and described warehouse to be allocated, according to institute
State the computing formula of characteristic, generate the described commodity to be allocated spy relevant to sales volume in warehouse to be allocated
Levy data.
Refer to Fig. 2, it is the tool of embodiment of the method step S101 of prediction inventory allocation ratio of the application
Body flow chart.Specifically, described acquisition commodity to be allocated are in the characteristic number relevant to sales volume in warehouse to be allocated
According to, comprise the steps:
Step S1011: in obtaining preset time range, described commodity to be allocated are at described warehouse to be allocated and sales volume
Relevant initial data.
Data have certain effect duration, data time then impact on inventory allocation scale prediction value the most remote
The least, in order to Accurate Prediction inventory allocation ratio, and reduce amount of calculation as far as possible, generally choose pre-
If the initial data in time range, for calculating the characteristic relevant to sales volume.In actual applications,
Can daily, week, the time range of the moon choose calculate initial data, such as, in the present embodiment,
Choose the commodity history 7 days in warehouse, within 14 days, 30 days, 60 days, to sell IPV/UV/ number of packages accountings etc. original
Data.The characteristic that initial data in the range of different time is formed is to the impact effect of final predictive value not
With, in actual applications, suitable characteristic can be chosen according to concrete demand.
Step S1013: the calculating according to the described initial data data relevant to sales volume and described characteristic is public
Formula, calculates and obtains the described characteristic relevant to sales volume.
After getting the initial data relevant to sales volume, according to the computing formula of default characteristic, meter
Calculate the various characteristics relevant to sales volume, in order to according to these characteristics, inventory allocation ratio is carried out
Prediction.Partial Feature data used by being set forth below in this enforcement and computing formula thereof:
1) preset the sale in preset time range of the default commodity in warehouse and click on ratio.
Described sale click ratio, refers to that the sale number of clicks presetting the default commodity in warehouse is default business
Total ratio sold in number of clicks of product.
2) preset the sale in preset time range of the default commodity in warehouse and click on buyer's ratio.
Described sale clicks on buyer's ratio, refers to that buyer's number is clicked in the sale presetting the default commodity in warehouse
Total at default commodity sells the ratio clicked in buyer's number.
3) preset the click of taking in preset time range of the default businessman belonging to default commodity in warehouse to compare
Example.
Described take click ratio, refer to preset taking a little of the default businessman belonging to default commodity in warehouse
Hit number of times, in the ratio always taken in number of clicks of the default businessman belonging to default commodity.
4) preset the click of taking in preset time range of the default businessman belonging to default commodity in warehouse to buy
Family's ratio.
Described taking clicks on buyer's ratio, refers to preset the bat of the default businessman belonging to default commodity in warehouse
Lower click buyer's number, in the ratio always taken in click buyer's number of the default businessman belonging to default commodity.
5) sale of the merchandise classification belonging to default commodity presetting businessman belonging to the default commodity in warehouse is preset
Ratio, sells ratio as second.
Described second sells ratio, refers to preset the default commodity presetting businessman belonging to the default commodity in warehouse
The sales volume of affiliated merchandise classification, is presetting the commodity belonging to the default commodity of businessman belonging to default commodity
Ratio in total sales volume of classification.
6) the sale ratio of the merchandise classification belonging to default commodity in default warehouse, sells ratio as the 3rd.
Described 3rd sells ratio, refers to preset the sale number of the merchandise classification belonging to default commodity in warehouse
Amount, the ratio in total sales volume of the merchandise classification belonging to default commodity.
7) commodity total sales volume in preset time range is preset.
8) preset the shopping cart in preset time range of the default commodity in warehouse and click on ratio.
Ratio clicked on by described shopping cart, refers to that the shopping cart number of clicks presetting the default commodity in warehouse is in advance
If the ratio in total shopping cart number of clicks of commodity.
9) preset the shopping cart in preset time range of the default commodity in warehouse and click on buyer's ratio.
Buyer's ratio clicked on by described shopping cart, refers to that buyer clicked on by the shopping cart presetting the default commodity in warehouse
Number clicks on the ratio in buyer's number at total shopping cart of default commodity.
10) the collection point in preset time range of the default businessman belonging to default commodity in warehouse is preset
Hit ratio.
Described collection clicks on ratio, refers to preset the collection number of clicks of the default businessman belonging to commodity,
Ratio in total collection number of clicks of the default businessman belonging to default commodity.
11) preset the collection in preset time range of the default businessman belonging to default commodity in warehouse to click on
Buyer's ratio.
Described collection clicks on buyer's ratio, refers to that the collection click presetting the default businessman belonging to commodity is bought
Family's number, the total collection in the default businessman belonging to default commodity clicks on the ratio in buyer's number.
12) predetermined ratio of the default commodity in default warehouse.
Described predetermined ratio, refers to total pre-at default commodity of the predetermined quantity presetting the default commodity in warehouse
Ratio in determined number.
13) the sale ratio of the default commodity in default warehouse, sells ratio as first.
Described first sells ratio, refers to that the sales volume presetting the default commodity in warehouse is at default commodity
Ratio in total sales volume.
Above-named is all some concrete characteristics, in actual applications, it is also possible to select other
The characteristic of form, different characteristics is all the change of concrete form, all without departing from the application's
Core, the most all within the protection domain of the application.
Step S103: according to described prediction characteristic and the forecast model that previously generates, treats described in generation
Distribution commodity are at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
Forecast model described herein is according to history feature data and corresponding inventory allocation ratio
(i.e. effective sale ratio) learns out, and the process therefore generating forecast model is a machine learning
Process, and the machine learning of the embodiment of the present application is the machine learning having supervision.Carry out the machine having supervision
Device learns, and the learning algorithm that can use includes regression algorithm.Regression algorithm belongs to the category of inductive learning,
So-called inductive learning refers to some examples according to certain concept, draws typically retouching of this concept by inductive reasoning
State.For the application of prediction commodity inventory allocation ratio in storehouse, it is possible to the regression algorithm of employing
Including linear regression, regression tree, iteration decision tree, or the weighted linear combination of default regression algorithm
Scheduling algorithm.The accuracy of the predictive value that algorithms of different generates is different, the computation complexity of algorithms of different the most not phase
With, in actual applications, according to concrete application demand, any one regression algorithm can be selected commodity
Inventory allocation ratio is predicted.
Refer to Fig. 3, its be the application prediction inventory allocation ratio embodiment of the method generate forecast model
Particular flow sheet.Specifically, forecast model described herein uses following steps to generate:
Step S301: obtain each commodity the history feature data relevant to sales volume in each warehouse and with institute
State the inventory allocation ratio that history feature data are corresponding--i.e. effective sale ratio, as training set.
In order to generate forecast model accurately, need to obtain substantial amounts of training data, it may be assumed that each commodity
The history feature data relevant to sales volume and the inventory allocation corresponding with history feature data in each warehouse
Ratio.The actual value of wherein corresponding with history feature data described inventory allocation ratio includes that commodity are in storehouse
Storehouse truly sells accounting.Commodity are used truly to sell accounting, as these commodity in this storehouse in warehouse
The actual value of the inventory allocation ratio that history feature data in storehouse are corresponding, its reason is: commodity are existed
Inventory allocation ratio in warehouse is just identical with effective sale ratio, thus reaches to obtain from training set learning
The higher effect of the accuracy of forecast model.
In actual applications, for reducing amount of calculation, ensure that the training data of foundation is abundant again simultaneously, need
Part sample is randomly selected as training data from all training datas.Such as, all training datas 100
Article ten thousand, i.e. there are 1,000,000 records, the information in every record to have recorded certain kinds of goods corresponding spy in certain warehouse
Levy, and the actual value of corresponding inventory allocation ratio, randomly select 100,000 records therein as instruction
Practice data.The training data form of the present embodiment is as shown in Table 1:
Table one, training data table
In table one, each characteristic relevant with sales volume is shown in each list in centre before last string, and last is classified as
The inventory allocation ratio corresponding with characteristic, i.e. effective sale ratio.
Step S303: by machine learning algorithm, obtain described forecast model from described training set learning.
In actual applications, generate forecast model according to training set, multiple concrete training algorithm can be used,
Including linear regression algorithm, RDT (Regression Decision Tree, regression tree) algorithm or GBDT
(Gradient Boosting Decision Tree, iteration decision tree) algorithm etc..By above-mentioned various algorithms,
Generate the forecast model for predicting inventory allocation ratio.Above-mentioned various different training algorithm is all concrete
The change of embodiment, all without departing from the core of the application, the most all within the protection domain of the application.
It is different by the accuracy of the predictive value of above-mentioned various different training algorithms generations.The application is real
Execute example provide method, characteristic is more, even as high as tens characteristics, thus characteristic with
Relation between predictive value is probably nonlinear, and linear regression algorithm is applicable to linear situation, therefore
The accuracy using the predictive value of linear regression algorithm generation is poor.RDT algorithm has the best characteristic,
Such as, training time complexity is relatively low, process ratio of prediction faster, (easily general easily shown by model
To decision tree make picture presentation out) etc..But, single decision tree is likely to occur the problem of over-fitting,
Although by certain methods, as beta pruning can reduce this situation, but inadequate.Over-fitting refers to
In order to make training set precision higher, acquire a lot " rule only set up in training set ", caused changing one
The current rule of data set is the most inapplicable.As long as the leaf node in fact allowing one tree is abundant, training set
Always can train 100% accuracy rate.Between training precision and available accuracy (or measuring accuracy), after
Person is only the target really needing to reach.
GBDT is again MART (Multiple Additive Regression Tree), is the decision-making of a kind of iteration
Tree algorithm, this algorithm is made up of many decision trees, and the conclusion of all trees adds up and does final result.It
It is considered as just the stronger algorithm of generalization ability (generalization) together with SVM at the beginning of being suggested.GBDT
May be used with nearly all regression problems (linear/non-linear), relative logistic regression is simply possible to use in linearly
Returning, the applicable surface of GBDT is the widest.The core of GBDT is that, every one tree be before institute
Have the residual error of tree conclusion sum, this residual error be exactly one add predictive value after can obtain the accumulation amount of actual value.
The final result of GBDT algorithm is to generate N (in the present embodiment, actual have more than a hundreds of) tree,
So can greatly reduce the shortcoming that single decision tree brings, although each in this hundreds of decision tree all
Very simple (for this single decision tree of C4.5), but they combine the most powerful.
Owing to GBDT is higher compared with the model prediction accuracy that RDT generates, thus the embodiment of the present application uses
GBDT algorithm generates forecast model.Application GBDT algorithm is given below in the present embodiment and builds forecast model
Process, and the prediction process corresponding with this algorithm.
Refer to Fig. 4, it is the tool of embodiment of the method step S303 of prediction inventory allocation ratio of the application
Body flow chart.
Step S3031: be respectively provided with the training of inventory allocation ratio for every training data in described training set
Use initial prediction.
Owing to the characteristic of the method institute foundation of the embodiment of the present application offer is more, even as high as tens spies
Levying data, therefore, the time complexity of the embodiment of the present application employing GBDT algorithm and computation complexity are the most relatively
High.In order to reduce time complexity and the computation complexity of training process so that forecast model is restrained as early as possible,
The training initial prediction of inventory allocation ratio it is respectively provided with for every training data in training set.In instruction
Practice stage and forecast period, owing to being provided with an initial value for predictive value in advance, thus the time of calculating general
To improving greatly.In the present embodiment, in order to reasonably arrange initial prediction, before current number of times
The meansigma methods of effective sale ratio of preset times as initial prediction, this is that one more meets reality
The initial prediction system of selection of application.
Step S3033: according to described training initial prediction and described training set, uses iteration decision tree to calculate
Method, generates described forecast model.
When being respectively provided with the training of inventory allocation ratio for every training data in described training set with initial pre-
After measured value, according to described training initial prediction and described training set, use iteration decision Tree algorithms, i.e.
Described forecast model can be generated.
According to training set, forecast model is trained, constructs N decision tree.Every decision tree of structure
Time, need to find the optimum split point of tree, tree is carried out the division of Y leaf node.Through great many of experiments
Show that the method that the embodiment of the present application provides carries out the division of 8 leaf nodes for every decision tree,
It is obtained in that accurate forecast model.Refer to Fig. 5, it is the side predicting inventory allocation ratio of the application
The schematic diagram of the forecast model that method embodiment generates.After completing step S303, training result is N tree.
The each tree built by GBDT algorithm is all a regression tree.Regression tree is more than one
Stage decision making process, it is not once to carry out decision-making by all features of training sample, but successively with each
Individual characteristic component carries out decision-making.The structure of regression tree generally comprises the steps: be 1) inside each
Node selects division rule;2) terminal note (Terminal Nodes) is determined.Regression tree at each node
(being not necessarily leaf node) all can obtain a predictive value, it may be assumed that inventory allocation ratio, and this predictive value is equal to
Belong to the meansigma methods of all commodity stocks allocation proportions of this node.During branch, each feature exhaustive is every
Individual threshold value looks for best cut-point, and weighing best standard is to minimize mean square deviation, it may be assumed that (the storehouse of each commodity
Deposit allocation proportion-prediction inventory allocation ratio) summation/N of ^2, or perhaps the forecast error of each commodity
Quadratic sum is divided by N.The most reliable branch foundation can be found by minimizing mean square deviation.Branch is until each
On leaf node, the inventory allocation ratio of commodity reaches default end condition (such as: in leaf node number
Limit), if the inventory allocation ratio of commodity is not unique, then with commodity all on this node on final leaf node
Average inventory allocation proportion is as the predictive value of the inventory allocation ratio of this leaf node.
Space hyperplane is actually carried out a kind of method divided by regression tree, every time segmentation time
Wait, all current space is divided into two so that each leaf node is the not phase of in space
The region handed over, carrying out decision-making when, can be according to the value of sample every one-dimensional characteristic data of input, a step
One step down, finally makes one (assuming there is N number of leaf node) that sample falls in N number of region.
In sum, it is required to find the optimum split point of tree during structure each tree, tree is carried out Y leaf
The division of node, i.e. all leaf nodes to present tree, carry out the process of following steps: 1) calculates each
(gain is loss function decrement, and loss decrement can be determined for the optimal dividing of leaf node and its gain
The difference of justice inventory allocation scale prediction value with inventory allocation ratio actual value for passing through the division of this leaf node
Quadratic sum);2) leaf node and the division points thereof that select gain maximum divide, and are drawn by training sample
Assign in child node;3) update inventory allocation scale prediction value, i.e. update certain commodity stock in certain warehouse and divide
Proportioning example predictive value.
Generating forecast model by above-mentioned steps S3031 and step S3033, this forecast model is for prediction rank
Section uses.At forecast period, the prediction characteristic got according to step S101 and above-mentioned employing GBDT
The forecast model that algorithm previously generates, to the described commodity to be allocated inventory allocation ratio in described warehouse to be allocated
Example is predicted.Such as, when to predict certain commodity Q inventory allocation ratio in two weeks futures of certain warehouse E
Predictive value time, first sort out the commodity Q current signature data at warehouse E, according to characteristic and pre-
The forecast model first generated, i.e. can get commodity Q inventory allocation ratio pre-in two weeks futures of warehouse E
Measured value, wherein current signature data are as shown in following table two:
Characteristic table is used in table two, prediction
Concrete, step S103 includes two steps:
Step S1031: the prediction of inventory allocation ratio is set in described warehouse to be allocated for described commodity to be allocated
Use initial prediction.
Step S1032: according to described prediction characteristic, prediction initial prediction and previously generate pre-
Survey model, generate the described commodity to be allocated predictive value in the inventory allocation ratio in described warehouse to be allocated.
Two concrete steps (step S1031 and step S1033) of above-mentioned forecast period and the two of the training stage
Individual concrete steps (step S3031 and step S3033) one_to_one corresponding, the two something in common repeats no more, no
It is that forecast period according to prediction characteristic, prediction initial prediction and previously generates with part
Forecast model, generates the described commodity to be allocated predictive value in the inventory allocation ratio in described warehouse to be allocated.
Refer to Fig. 6, its be the application prediction inventory allocation ratio embodiment of the method prediction process signal
Figure.In the present embodiment, according to kinds of goods Q at the actual value of the first six week inventory allocation ratio of warehouse E, count
It is 20% that calculation obtains kinds of goods Q at the initial prediction of the warehouse E inventory allocation ratio of following two weeks, according in advance
Survey characteristic, uses N the tree that the training stage generates, obtains final predictive value: commodity Q is in warehouse
Predictive value=20% (initial prediction) of the E inventory allocation ratio of following two weeks be+7% (one tree
Gain)-3% (gain of second tree)+2% (gain of the 3rd tree)-...=18%.
The method of the prediction inventory allocation ratio that the embodiment of the present application provides, application machine learning thought pair and pin
The data that amount is relevant are analyzed, and use GBDT algorithm set up forecast model and use it to be predicted, band
Return checking and show that this forecast model has degree of precision.After machine acquistion data rule, utilize prediction mould
Type, carries out the analysis of knowledge acquisition to the model after acquistion knowledge, draws by having that machine-learning process obtains
The inventory allocation scale prediction value of meaning, provides objective basis for decision-making, has the strongest practicality.
In actual applications, in order to improve forecasting accuracy, it is also possible to multiple regression algorithms are generated respectively
The combination of various concrete forecast models is as final forecast model.Concrete, pass through machine described in employing
Learning algorithm, obtains described forecast model from described training set learning, in the following way: according to presetting
Various machine learning algorithms, obtain from described training set learning and described default various machine learning calculated
The forecast model that method is the most corresponding.
The various machine learning algorithms preset described in the embodiment of the present application can be various regression algorithm, specifically
Including linear regression algorithm, regression tree algorithm, iteration decision Tree algorithms or random forests algorithm scheduling algorithm.
In actual applications, learning algorithm can be selected according to concrete application demand, such as: select linear regression to calculate
Method and iteration decision Tree algorithms.
The multiple concrete forecast model generated with the above-mentioned training stage is corresponding, at forecast period, and described
According to described prediction characteristic and the forecast model previously generated, generate described commodity to be allocated and treat described
The predictive value of the inventory allocation ratio of distribution depot, comprises the steps: 1) to described prediction characteristic,
Each forecast model being respectively adopted study acquisition carries out the prediction of described inventory allocation ratio, it is thus achieved that each is pre-
Survey the predictive value of the described inventory allocation ratio that model prediction goes out;2) basis dopes for each forecast model
The weight that predictive value is preset respectively, the described inventory allocation ratio that each forecast model described is doped pre-
Measured value is weighted average computation, as the described commodity to be allocated inventory allocation ratio in described warehouse to be allocated
The predictive value of example.
1) to described prediction characteristic, each forecast model being respectively adopted study acquisition carries out described storehouse
Deposit the prediction of allocation proportion, it is thus achieved that the predictive value of the described inventory allocation ratio that each forecast model dopes.
When according to the various machine learning algorithms preset, obtain and the various machines preset from training set learning
After the forecast model that learning algorithm is the most corresponding, at forecast period, it is possible to be respectively adopted each of study acquisition
Individual forecast model carries out the prediction of described inventory allocation ratio, thus obtains the institute that each forecast model dopes
State the predictive value of inventory allocation ratio.
2) weight preset respectively according to the predictive value doped for each forecast model, to each prediction described
The predictive value of the described inventory allocation ratio that model prediction goes out is weighted average computation, as described to be allocated
Commodity are at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
Owing to various different machine learning algorithms are respectively arranged with pluses and minuses, the precision of the predictive value of generation also differs,
Preset weight respectively for this predictive value that can dope for each forecast model in advance, then each is predicted
The predictive value of the inventory allocation ratio that model prediction goes out is weighted average computation, is treating as commodity to be allocated
The final predictive value of the inventory allocation ratio of distribution depot.Simply, can be that each forecast model dopes
Predictive value preset identical weight.
It is in the above-described embodiment, it is provided that a kind of method predicting inventory allocation ratio, corresponding,
The application also provides for a kind of device predicting inventory allocation ratio.This device is and above-mentioned prediction inventory allocation ratio
The embodiment of the method for example is corresponding.
Refer to Fig. 7, it is the schematic diagram of device embodiment of prediction inventory allocation ratio of the application.Due to
Device embodiment is substantially similar to embodiment of the method, so describing fairly simple, relevant part sees method
The part of embodiment illustrates.Device embodiment described below is only schematically.
A kind of device predicting inventory allocation ratio of the present embodiment, including:
Acquiring unit 101, for obtaining the commodity to be allocated characteristic number relevant to sales volume in warehouse to be allocated
According to, as prediction characteristic;
Predicting unit 103, for according to described prediction characteristic and the forecast model previously generated, generating
Described commodity to be allocated are at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
Refer to Fig. 8, it is the concrete schematic diagram of device embodiment of prediction inventory allocation ratio of the application.
Optionally, also include:
Training unit 201, is used for generating described forecast model.
Optionally, described training unit 201 includes:
Obtain subelement 2011, for obtaining each commodity history feature relevant to sales volume in each warehouse
Data and the inventory allocation ratio corresponding with described history feature data--i.e. effective sale ratio, as training
Collection;
Study subelement 2013, for by machine learning algorithm, obtaining described from described training set learning
Forecast model.
Optionally, described study subelement 2013 includes:
Initial value arranges subelement, divides for being respectively provided with stock for every training data in described training set
The training initial prediction of proportioning example;
Generate subelement, for according to described training initial prediction and described training set, use iteration certainly
Plan tree algorithm, generates described forecast model.
Optionally, the effective sale of the preset times before described training initial prediction uses current number of times
The meansigma methods of ratio.
Optionally, described predicting unit 103 includes:
Initial value arranges subelement, divides for arranging stock for described commodity to be allocated in described warehouse to be allocated
The prediction initial prediction of proportioning example;
Computation subunit, for described according to described prediction characteristic, prediction initial prediction with pre-
The forecast model first generated, generates the described commodity to be allocated inventory allocation ratio in described warehouse to be allocated
Predictive value.
Optionally, described obtain described forecast model by machine learning algorithm from described training set learning,
In the following way:
According to default various machine learning algorithms, obtain default each with described from described training set learning
Plant the forecast model that machine learning algorithm is the most corresponding.
Optionally, described predicting unit 103 includes:
Prediction subelement, for described prediction characteristic, being respectively adopted each prediction that study obtains
Model carries out the prediction of described inventory allocation ratio, it is thus achieved that the described inventory allocation that each forecast model dopes
The predictive value of ratio;
Computation subunit, for the weight preset respectively according to the predictive value doped for each forecast model,
The predictive value of the described inventory allocation ratio doping each forecast model described is weighted average computation,
As described commodity to be allocated at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
Optionally, described acquiring unit 101 includes:
Obtaining subelement, in being used for obtaining preset time range, described commodity to be allocated are in described warehouse to be allocated
The initial data relevant to sales volume;
Computation subunit, for according to the described initial data relevant to sales volume and the calculating of described characteristic
Formula, calculates and obtains the described characteristic relevant to sales volume.
In the above-described embodiment, it is provided that a kind of method predicting inventory allocation ratio and corresponding intrument, originally
Application also provides for a kind of electronic equipment.This equipment is corresponding with the embodiment of said method and corresponding intrument.
Refer to Fig. 9, it is the flow chart of electronic equipment embodiment of the application.Owing to apparatus embodiments is basic
Similar in appearance to embodiment of the method, so describing fairly simple, relevant part sees the part of embodiment of the method and says
Bright.Apparatus embodiments described below is only schematically.
The application also provides for a kind of electronic equipment, including: display 901;Processor 902;And memorizer
903, for the device of Storage Estimation inventory allocation ratio, the device of described prediction inventory allocation ratio is described
When processor 902 performs, comprise the steps: to obtain commodity to be allocated warehouse to be allocated with sales volume phase
The characteristic closed, as prediction characteristic;According to described prediction characteristic with previously generate
Forecast model, generates the described commodity to be allocated predictive value in the inventory allocation ratio in described warehouse to be allocated.
The prediction method of inventory allocation ratio that the application provides, device and electronic equipment, treat point by obtaining
Join the commodity characteristic relevant to sales volume in warehouse to be allocated, and according to the characteristic obtained and in advance
The forecast model generated, generates the commodity to be allocated predictive value in the inventory allocation ratio in warehouse to be allocated.By
The method provided in the application is historical sales behavior based on commodity and issuable purchasing demand in the future
Assess the commodity to be allocated allocation proportion in warehouse to be allocated etc. combined factors, this improves inventory allocation
The prediction accuracy of ratio, it is to avoid too much rely on manpower and carry out inventory allocation, thus reduce Out of Stock
Rate and dispensing timeliness, improve stock's turnover level and the effect of consumer experience.
Although the application is open as above with preferred embodiment, but it is not for limiting the application, Ren Heben
Skilled person, without departing from spirit and scope, can make possible variation and amendment,
Therefore the protection domain of the application should be defined in the range of standard with the application claim.
In a typical configuration, calculating equipment includes one or more processor (CPU), input/output
Interface, network interface and internal memory.
Internal memory potentially includes the volatile memory in computer-readable medium, random access memory
(RAM) and/or the form such as Nonvolatile memory, such as read only memory (ROM) or flash memory (flash RAM).
Internal memory is the example of computer-readable medium.
1, computer-readable medium includes that permanent and non-permanent, removable and non-removable media can be by
Any method or technology realize information storage.Information can be computer-readable instruction, data structure, journey
The module of sequence or other data.The example of the storage medium of computer includes, but are not limited to phase transition internal memory
(PRAM), static RAM (SRAM), dynamic random access memory (DRAM), its
The random access memory (RAM) of his type, read only memory (ROM), electrically erasable is read-only deposits
Reservoir (EEPROM), fast flash memory bank or other memory techniques, read-only optical disc read only memory
(CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tape, tape magnetic magnetic
Disk storage or other magnetic storage apparatus or any other non-transmission medium, can be used for storage can be set by calculating
The standby information accessed.According to defining herein, computer-readable medium does not include non-temporary computer-readable matchmaker
Body (transitory media), such as data signal and the carrier wave of modulation.
2, it will be understood by those skilled in the art that embodiments herein can be provided as method, system or computer
Program product.Therefore, the application can use complete hardware embodiment, complete software implementation or combine software
Form with the embodiment of hardware aspect.And, the application can use and wherein include meter one or more
The computer-usable storage medium of calculation machine usable program code (include but not limited to disk memory,
CD-ROM, optical memory etc.) form of the upper computer program implemented.
Claims (23)
1. the method predicting inventory allocation ratio, it is characterised in that including:
Obtain the commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated, as prediction feature
Data;
According to described prediction characteristic and the forecast model previously generated, generate described commodity to be allocated and exist
The predictive value of the inventory allocation ratio in described warehouse to be allocated.
The method of prediction inventory allocation ratio the most according to claim 1, it is characterised in that described pre-
Surveying model uses following steps to generate:
Obtain each commodity the history feature data relevant to sales volume in each warehouse and with described history feature
The inventory allocation ratio that data are corresponding--i.e. effective sale ratio, as training set;
By machine learning algorithm, obtain described forecast model from described training set learning.
The method of prediction inventory allocation ratio the most according to claim 2, it is characterised in that described machine
Device learning algorithm includes regression algorithm.
The method of prediction inventory allocation ratio the most according to claim 3, it is characterised in that described time
Reduction method includes linear regression algorithm, regression tree algorithm, iteration decision Tree algorithms or random forests algorithm.
The method of prediction inventory allocation ratio the most according to claim 2, it is characterised in that described logical
Cross machine learning algorithm, obtain described forecast model from described training set learning, including:
The training initial predicted of inventory allocation ratio it is respectively provided with for every training data in described training set
Value;
According to described training initial prediction and described training set, use iteration decision Tree algorithms, generate institute
State forecast model.
The method of prediction inventory allocation ratio the most according to claim 5, it is characterised in that described instruction
White silk initial prediction uses the meansigma methods of the effective sale ratio of the preset times before current number of times.
The method of prediction inventory allocation ratio the most according to claim 5, it is characterised in that described
According to described prediction characteristic and the forecast model previously generated, generate described commodity to be allocated and treat described
The predictive value of the inventory allocation ratio of distribution depot, including:
The prediction initial predicted of inventory allocation ratio is set in described warehouse to be allocated for described commodity to be allocated
Value;
According to described prediction characteristic, prediction initial prediction and the forecast model previously generated, raw
Become described commodity to be allocated at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
The method of prediction inventory allocation ratio the most according to claim 7, it is characterised in that described pre-
Survey initial prediction uses the meansigma methods of the effective sale ratio of the preset times before current number of times.
The method of prediction inventory allocation ratio the most according to claim 2, it is characterised in that described logical
Cross machine learning algorithm, obtain described forecast model from described training set learning, in the following way:
According to default various machine learning algorithms, obtain default each with described from described training set learning
Plant the forecast model that machine learning algorithm is the most corresponding.
The method of prediction inventory allocation ratio the most according to claim 9, it is characterised in that described
According to described prediction characteristic and the forecast model previously generated, generate described commodity to be allocated described
The predictive value of the inventory allocation ratio in warehouse to be allocated, including:
To described prediction characteristic, each forecast model being respectively adopted study acquisition carries out described stock
The prediction of allocation proportion, it is thus achieved that the predictive value of the described inventory allocation ratio that each forecast model dopes;
The weight preset respectively according to the predictive value doped for each forecast model, to each prediction mould described
The predictive value of the described inventory allocation ratio that type dopes is weighted average computation, as described business to be allocated
Product are at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
The method of 11. prediction inventory allocation ratios according to claim 1, it is characterised in that described in obtain
Take the commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated, including:
Obtain original relevant to sales volume in described warehouse to be allocated of described commodity to be allocated in preset time range
Data;
According to the described initial data relevant to sales volume and the computing formula of described characteristic, calculate and obtain institute
State the characteristic relevant to sales volume.
12. according to the method predicting inventory allocation ratio described in claim 1-11 any one, its feature
Being, the described characteristic relevant to sales volume includes at least one of following data:
Preset the sale in preset time range of the default commodity in warehouse and click on ratio;Described sale is clicked on
Ratio, refers to that the sale number of clicks presetting the default commodity in warehouse sells click time the total of default commodity
Ratio in number;
Preset the sale in preset time range of the default commodity in warehouse and click on buyer's ratio;Described sale
Click on buyer's ratio, refer to that the selling of default commodity presetting in warehouse clicks on buyer's number at default commodity
Total ratio sold in click buyer's number;
Preset the click of taking in preset time range of the default businessman belonging to default commodity in warehouse to compare
Example;Described take click ratio, refer to preset taking a little of the default businessman belonging to default commodity in warehouse
Hit number of times, in the ratio always taken in number of clicks of the default businessman belonging to default commodity;
Preset the taking in preset time range of the default businessman belonging to default commodity in warehouse and click on buyer
Ratio;Described take click buyer's ratio, refer to preset the default businessman belonging to default commodity in warehouse
Take click buyer's number, at the ratio always taken in click buyer's number of the default businessman belonging to default commodity
Example;
Preset the sale ratio of the merchandise classification belonging to default commodity presetting businessman belonging to the default commodity in warehouse
Example, sells ratio as second;Described second sells ratio, refers to preset belonging to the default commodity in warehouse
Preset the sales volume of the merchandise classification belonging to default commodity of businessman, belonging to default commodity, preset businessman
Preset the ratio in total sales volume of the merchandise classification belonging to commodity;
The sale ratio of the merchandise classification belonging to default commodity in default warehouse, sells ratio as the 3rd;
Described 3rd sells ratio, refers to preset the sales volume of the merchandise classification belonging to default commodity in warehouse,
Ratio in total sales volume of the merchandise classification belonging to default commodity;
Preset the commodity total sales volume in preset time range;
Preset the shopping cart in preset time range of the default commodity in warehouse and click on ratio;Described shopping cart
Click ratio, refers to the shopping cart number of clicks presetting the default commodity in warehouse total shopping at default commodity
Ratio in car number of clicks;
Preset the shopping cart in preset time range of the default commodity in warehouse and click on buyer's ratio;Described purchase
Buyer's ratio clicked on by thing car, refers to that the shopping cart presetting the default commodity in warehouse clicks on buyer's number default
The ratio in buyer's number clicked on by total shopping cart of commodity;
Preset the collection in preset time range of the default businessman belonging to default commodity in warehouse and click on ratio
Example;Described collection clicks on ratio, refers to preset the collection number of clicks of the default businessman belonging to commodity,
Ratio in total collection number of clicks of the default businessman belonging to default commodity;
Preset the collection click in preset time range of the default businessman belonging to default commodity in warehouse to buy
Family's ratio;Described collection clicks on buyer's ratio, refers to preset the collection point of the default businessman belonging to commodity
Hitting buyer's number, the total collection in the default businessman belonging to default commodity clicks on the ratio in buyer's number;
The predetermined ratio of the default commodity in default warehouse;Described predetermined ratio, refer to preset in warehouse is pre-
If the ratio that the predetermined quantity of commodity is in total predetermined quantity of default commodity.
The method of 13. prediction inventory allocation ratios according to claim 12, it is characterised in that described
The characteristic relevant to sales volume also includes:
The sale ratio of the default commodity in default warehouse, sells ratio as first;Described first sells ratio
Example, refers to the sales volume presetting the default commodity in warehouse ratio in total sales volume of default commodity.
14. 1 kinds of devices predicting inventory allocation ratio, it is characterised in that including:
Acquiring unit, for obtaining the commodity to be allocated characteristic relevant to sales volume in warehouse to be allocated,
As prediction characteristic;
Predicting unit, for according to described prediction characteristic and the forecast model previously generated, generating institute
State the commodity to be allocated predictive value in the inventory allocation ratio in described warehouse to be allocated.
The device of 15. prediction inventory allocation ratios according to claim 14, it is characterised in that also wrap
Include:
Training unit, is used for generating described forecast model.
The device of 16. prediction inventory allocation ratios according to claim 15, it is characterised in that described
Training unit includes:
Obtain subelement, for obtaining each commodity history feature data relevant to sales volume in each warehouse
And the inventory allocation ratio corresponding with described history feature data--i.e. effective sale ratio, as training set;
Study subelement, for by machine learning algorithm, obtains described prediction from described training set learning
Model.
The device of 17. prediction inventory allocation ratios according to claim 16, it is characterised in that described
Study subelement includes:
Initial value arranges subelement, divides for being respectively provided with stock for every training data in described training set
The training initial prediction of proportioning example;
Generate subelement, for according to described training initial prediction and described training set, use iteration certainly
Plan tree algorithm, generates described forecast model.
The device of 18. prediction inventory allocation ratios according to claim 17, it is characterised in that described
Training initial prediction uses the meansigma methods of the effective sale ratio of the preset times before current number of times.
The device of 19. prediction inventory allocation ratios according to claim 17, it is characterised in that described
Predicting unit includes:
Initial value arranges subelement, divides for arranging stock for described commodity to be allocated in described warehouse to be allocated
The prediction initial prediction of proportioning example;
Computation subunit, for according to described prediction characteristic, prediction initial prediction and pre-Mr.
The forecast model become, generates the prediction in the inventory allocation ratio in described warehouse to be allocated of the described commodity to be allocated
Value.
The device of 20. prediction inventory allocation ratios according to claim 16, it is characterised in that described
By machine learning algorithm, obtain described forecast model from described training set learning, in the following way:
According to default various machine learning algorithms, obtain default each with described from described training set learning
Plant the forecast model that machine learning algorithm is the most corresponding.
The device of 21. prediction inventory allocation ratios according to claim 20, it is characterised in that described
Predicting unit includes:
Prediction subelement, for described prediction characteristic, being respectively adopted each prediction that study obtains
Model carries out the prediction of described inventory allocation ratio, it is thus achieved that the described inventory allocation that each forecast model dopes
The predictive value of ratio;
Computation subunit, for the weight preset respectively according to the predictive value doped for each forecast model,
The predictive value of the described inventory allocation ratio doping each forecast model described is weighted average computation,
As described commodity to be allocated at the predictive value of the inventory allocation ratio in described warehouse to be allocated.
The device of 22. prediction inventory allocation ratios according to claim 14, it is characterised in that described
Acquiring unit includes:
Obtaining subelement, in being used for obtaining preset time range, described commodity to be allocated are in described warehouse to be allocated
The initial data relevant to sales volume;
Computation subunit, for according to the described initial data relevant to sales volume and the calculating of described characteristic
Formula, calculates and obtains the described characteristic relevant to sales volume.
23. 1 kinds of electronic equipments, it is characterised in that including:
Display;
Processor;And
Memorizer, for the device of Storage Estimation inventory allocation ratio, the dress of described prediction inventory allocation ratio
Put when being performed by described processor, comprise the steps: to obtain commodity to be allocated warehouse to be allocated with pin
The characteristic that amount is relevant, as prediction characteristic;According to described prediction characteristic and pre-Mr.
The forecast model become, generates the prediction in the inventory allocation ratio in described warehouse to be allocated of the described commodity to be allocated
Value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510194457.5A CN106156880A (en) | 2015-04-22 | 2015-04-22 | A kind of predict the method for inventory allocation ratio, device and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510194457.5A CN106156880A (en) | 2015-04-22 | 2015-04-22 | A kind of predict the method for inventory allocation ratio, device and electronic equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106156880A true CN106156880A (en) | 2016-11-23 |
Family
ID=57346435
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510194457.5A Pending CN106156880A (en) | 2015-04-22 | 2015-04-22 | A kind of predict the method for inventory allocation ratio, device and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106156880A (en) |
Cited By (43)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106897795A (en) * | 2017-02-17 | 2017-06-27 | 联想(北京)有限公司 | A kind of inventory forecast method and device |
CN107016571A (en) * | 2017-03-31 | 2017-08-04 | 北京百分点信息科技有限公司 | Data predication method and its system |
CN107067306A (en) * | 2017-04-17 | 2017-08-18 | 王晩艳 | Shopping at network method and system |
CN107230035A (en) * | 2017-06-29 | 2017-10-03 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
CN107464077A (en) * | 2017-07-21 | 2017-12-12 | 上海杉互健康科技有限公司 | A kind of trial drug inventory management system for applying to research center |
CN107563702A (en) * | 2017-09-14 | 2018-01-09 | 北京京东尚科信息技术有限公司 | Commodity storage concocting method, device and storage medium |
CN107886279A (en) * | 2017-11-29 | 2018-04-06 | 商派软件有限公司 | A kind of full channel inventory allocation method based on businessman stock |
CN108108933A (en) * | 2017-12-01 | 2018-06-01 | 中国联合网络通信集团有限公司 | Position distribution method of storing in a warehouse and device |
CN108133402A (en) * | 2017-12-05 | 2018-06-08 | 天脉聚源(北京)科技有限公司 | Commodity choosing method and device |
CN108536650A (en) * | 2018-04-03 | 2018-09-14 | 北京京东尚科信息技术有限公司 | Generate the method and apparatus that gradient promotes tree-model |
CN108629436A (en) * | 2017-03-15 | 2018-10-09 | 阿里巴巴集团控股有限公司 | A kind of method and electronic equipment of estimation warehouse picking ability |
CN108830659A (en) * | 2018-06-21 | 2018-11-16 | 杭州骑轻尘信息技术有限公司 | Vehicle relation between supply and demand appraisal procedure, device and electronic equipment |
CN108874805A (en) * | 2017-05-09 | 2018-11-23 | 腾讯科技(北京)有限公司 | The treating method and apparatus of data |
CN108932636A (en) * | 2018-07-05 | 2018-12-04 | 上海信颐信息技术有限公司 | A kind of commodity retail prediction technique and equipment |
CN108985807A (en) * | 2017-05-31 | 2018-12-11 | 北京京东尚科信息技术有限公司 | The method and apparatus for determining article characteristics type |
CN109064203A (en) * | 2018-06-12 | 2018-12-21 | 佛山欧神诺陶瓷有限公司 | A kind of BI data statistics method for early warning |
CN109064274A (en) * | 2018-07-24 | 2018-12-21 | 方骥 | A kind of wisdom electric business page shortage of goods management method and system based on machine learning |
CN109360022A (en) * | 2018-10-15 | 2019-02-19 | 广东工业大学 | A kind of market Sales Volume of Commodity prediction technique, device and equipment based on data mining |
CN109509030A (en) * | 2018-11-15 | 2019-03-22 | 北京旷视科技有限公司 | Method for Sales Forecast method and its training method of model, device and electronic system |
CN109754199A (en) * | 2017-11-07 | 2019-05-14 | 北京京东尚科信息技术有限公司 | Information output method and device |
CN109784979A (en) * | 2018-12-19 | 2019-05-21 | 重庆邮电大学 | A kind of supply chain needing forecasting method of big data driving |
CN109840730A (en) * | 2017-11-29 | 2019-06-04 | 北京京东尚科信息技术有限公司 | Method and device for data prediction |
CN109961306A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | A kind of inventory allocation method and apparatus of article |
CN109961247A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | A kind of generation method and device of article storage information |
CN109978421A (en) * | 2017-12-28 | 2019-07-05 | 北京京东尚科信息技术有限公司 | Information output method and device |
CN110046861A (en) * | 2019-04-24 | 2019-07-23 | 北京百度网讯科技有限公司 | Inventory management method and device, electronic equipment, computer-readable medium |
CN110489719A (en) * | 2019-07-31 | 2019-11-22 | 天津大学 | Wind speed forecasting method based on DATA PROCESSING IN ENSEMBLE PREDICTION SYSTEM data |
CN110619407A (en) * | 2018-06-19 | 2019-12-27 | 北京京东尚科信息技术有限公司 | Object sales prediction method and system, electronic device, and storage medium |
CN111062661A (en) * | 2019-12-10 | 2020-04-24 | 北京云杉信息技术有限公司 | Sales prediction-based storage capacity allocation method and device |
CN111125181A (en) * | 2018-10-31 | 2020-05-08 | 北京国双科技有限公司 | Method and device for obtaining age ratio, machine-readable storage medium and processor |
CN111367927A (en) * | 2020-03-02 | 2020-07-03 | 壹车宜家信息科技有限公司 | Automobile consumption data management method based on big data analysis |
CN111598495A (en) * | 2019-02-20 | 2020-08-28 | 网易(杭州)网络有限公司 | Output unit amount prediction method, medium, device and computing equipment |
CN112085442A (en) * | 2020-08-27 | 2020-12-15 | 北京每日优鲜电子商务有限公司 | Device operation method, device, electronic device and computer readable medium |
CN112330037A (en) * | 2020-11-11 | 2021-02-05 | 天津汇商共达科技有限责任公司 | Method and device for predicting inventory proportion of new product and server |
CN112597745A (en) * | 2020-12-23 | 2021-04-02 | 深圳赛安特技术服务有限公司 | Intelligent chart generation method and device, computer system and readable storage medium |
CN113065882A (en) * | 2020-01-02 | 2021-07-02 | 阿里巴巴集团控股有限公司 | Commodity processing method and device and electronic equipment |
CN113128932A (en) * | 2021-04-16 | 2021-07-16 | 北京京东振世信息技术有限公司 | Warehouse stock processing method and device, storage medium and electronic equipment |
CN113139767A (en) * | 2020-01-20 | 2021-07-20 | 菜鸟智能物流控股有限公司 | Logistics object allocation method and device, electronic equipment and computer-readable storage medium |
CN113298546A (en) * | 2020-05-29 | 2021-08-24 | 阿里巴巴集团控股有限公司 | Sales prediction method and device, and commodity processing method and device |
CN113298552A (en) * | 2020-12-02 | 2021-08-24 | 阿里巴巴集团控股有限公司 | Data processing method, server and storage medium |
CN113657822A (en) * | 2021-07-20 | 2021-11-16 | 中国邮电器材集团有限公司 | Method and device for distributing commodity quota |
TWI759921B (en) * | 2019-12-06 | 2022-04-01 | 南韓商韓領有限公司 | Supply chain management system and supply chain management method |
WO2022199383A1 (en) * | 2021-03-26 | 2022-09-29 | 北京京东拓先科技有限公司 | Article replenishment method and device, and storage medium |
-
2015
- 2015-04-22 CN CN201510194457.5A patent/CN106156880A/en active Pending
Cited By (57)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106897795A (en) * | 2017-02-17 | 2017-06-27 | 联想(北京)有限公司 | A kind of inventory forecast method and device |
CN108629436A (en) * | 2017-03-15 | 2018-10-09 | 阿里巴巴集团控股有限公司 | A kind of method and electronic equipment of estimation warehouse picking ability |
CN108629436B (en) * | 2017-03-15 | 2022-04-12 | 阿里巴巴集团控股有限公司 | Method and electronic equipment for estimating warehouse goods picking capacity |
CN107016571A (en) * | 2017-03-31 | 2017-08-04 | 北京百分点信息科技有限公司 | Data predication method and its system |
CN107067306A (en) * | 2017-04-17 | 2017-08-18 | 王晩艳 | Shopping at network method and system |
CN108874805A (en) * | 2017-05-09 | 2018-11-23 | 腾讯科技(北京)有限公司 | The treating method and apparatus of data |
CN108874805B (en) * | 2017-05-09 | 2022-05-20 | 腾讯科技(北京)有限公司 | Data processing method and device |
CN108985807A (en) * | 2017-05-31 | 2018-12-11 | 北京京东尚科信息技术有限公司 | The method and apparatus for determining article characteristics type |
CN107230035A (en) * | 2017-06-29 | 2017-10-03 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
CN107230035B (en) * | 2017-06-29 | 2021-03-30 | 北京京东尚科信息技术有限公司 | Information pushing method and device |
CN107464077A (en) * | 2017-07-21 | 2017-12-12 | 上海杉互健康科技有限公司 | A kind of trial drug inventory management system for applying to research center |
CN107464077B (en) * | 2017-07-21 | 2020-11-03 | 上海杉互健康科技有限公司 | Test drug inventory management system applied to research center |
CN107563702A (en) * | 2017-09-14 | 2018-01-09 | 北京京东尚科信息技术有限公司 | Commodity storage concocting method, device and storage medium |
CN109754199B (en) * | 2017-11-07 | 2021-07-20 | 北京京东尚科信息技术有限公司 | Information output method and device |
CN109754199A (en) * | 2017-11-07 | 2019-05-14 | 北京京东尚科信息技术有限公司 | Information output method and device |
CN109840730A (en) * | 2017-11-29 | 2019-06-04 | 北京京东尚科信息技术有限公司 | Method and device for data prediction |
CN107886279A (en) * | 2017-11-29 | 2018-04-06 | 商派软件有限公司 | A kind of full channel inventory allocation method based on businessman stock |
CN108108933A (en) * | 2017-12-01 | 2018-06-01 | 中国联合网络通信集团有限公司 | Position distribution method of storing in a warehouse and device |
CN108108933B (en) * | 2017-12-01 | 2022-02-22 | 中国联合网络通信集团有限公司 | Storage position distribution method and device |
CN108133402A (en) * | 2017-12-05 | 2018-06-08 | 天脉聚源(北京)科技有限公司 | Commodity choosing method and device |
CN109961306B (en) * | 2017-12-25 | 2022-04-12 | 北京京东尚科信息技术有限公司 | Method and device for distributing inventory of articles |
CN109961306A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | A kind of inventory allocation method and apparatus of article |
CN109961247A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | A kind of generation method and device of article storage information |
CN109978421A (en) * | 2017-12-28 | 2019-07-05 | 北京京东尚科信息技术有限公司 | Information output method and device |
CN109978421B (en) * | 2017-12-28 | 2021-04-30 | 北京京东尚科信息技术有限公司 | Information output method and device |
CN108536650B (en) * | 2018-04-03 | 2022-04-26 | 北京京东尚科信息技术有限公司 | Method and device for generating gradient lifting tree model |
CN108536650A (en) * | 2018-04-03 | 2018-09-14 | 北京京东尚科信息技术有限公司 | Generate the method and apparatus that gradient promotes tree-model |
CN109064203A (en) * | 2018-06-12 | 2018-12-21 | 佛山欧神诺陶瓷有限公司 | A kind of BI data statistics method for early warning |
CN110619407B (en) * | 2018-06-19 | 2024-04-09 | 北京京东尚科信息技术有限公司 | Object sales prediction method and system, electronic device and storage medium |
CN110619407A (en) * | 2018-06-19 | 2019-12-27 | 北京京东尚科信息技术有限公司 | Object sales prediction method and system, electronic device, and storage medium |
CN108830659A (en) * | 2018-06-21 | 2018-11-16 | 杭州骑轻尘信息技术有限公司 | Vehicle relation between supply and demand appraisal procedure, device and electronic equipment |
CN108932636A (en) * | 2018-07-05 | 2018-12-04 | 上海信颐信息技术有限公司 | A kind of commodity retail prediction technique and equipment |
CN109064274A (en) * | 2018-07-24 | 2018-12-21 | 方骥 | A kind of wisdom electric business page shortage of goods management method and system based on machine learning |
CN109064274B (en) * | 2018-07-24 | 2021-09-21 | 方骥 | Intelligent e-commerce page out-of-stock management method and system based on machine learning |
CN109360022A (en) * | 2018-10-15 | 2019-02-19 | 广东工业大学 | A kind of market Sales Volume of Commodity prediction technique, device and equipment based on data mining |
CN111125181A (en) * | 2018-10-31 | 2020-05-08 | 北京国双科技有限公司 | Method and device for obtaining age ratio, machine-readable storage medium and processor |
CN109509030A (en) * | 2018-11-15 | 2019-03-22 | 北京旷视科技有限公司 | Method for Sales Forecast method and its training method of model, device and electronic system |
CN109784979A (en) * | 2018-12-19 | 2019-05-21 | 重庆邮电大学 | A kind of supply chain needing forecasting method of big data driving |
CN111598495B (en) * | 2019-02-20 | 2023-07-25 | 网易(杭州)网络有限公司 | Method, medium, device and computing equipment for predicting quantity of orders |
CN111598495A (en) * | 2019-02-20 | 2020-08-28 | 网易(杭州)网络有限公司 | Output unit amount prediction method, medium, device and computing equipment |
CN110046861A (en) * | 2019-04-24 | 2019-07-23 | 北京百度网讯科技有限公司 | Inventory management method and device, electronic equipment, computer-readable medium |
CN110489719A (en) * | 2019-07-31 | 2019-11-22 | 天津大学 | Wind speed forecasting method based on DATA PROCESSING IN ENSEMBLE PREDICTION SYSTEM data |
TWI759921B (en) * | 2019-12-06 | 2022-04-01 | 南韓商韓領有限公司 | Supply chain management system and supply chain management method |
CN111062661A (en) * | 2019-12-10 | 2020-04-24 | 北京云杉信息技术有限公司 | Sales prediction-based storage capacity allocation method and device |
CN113065882A (en) * | 2020-01-02 | 2021-07-02 | 阿里巴巴集团控股有限公司 | Commodity processing method and device and electronic equipment |
CN113139767A (en) * | 2020-01-20 | 2021-07-20 | 菜鸟智能物流控股有限公司 | Logistics object allocation method and device, electronic equipment and computer-readable storage medium |
CN111367927A (en) * | 2020-03-02 | 2020-07-03 | 壹车宜家信息科技有限公司 | Automobile consumption data management method based on big data analysis |
CN113298546A (en) * | 2020-05-29 | 2021-08-24 | 阿里巴巴集团控股有限公司 | Sales prediction method and device, and commodity processing method and device |
CN112085442A (en) * | 2020-08-27 | 2020-12-15 | 北京每日优鲜电子商务有限公司 | Device operation method, device, electronic device and computer readable medium |
CN112085442B (en) * | 2020-08-27 | 2023-12-22 | 太原市万尺云数智科技有限公司 | Device operation method, apparatus, electronic device, and computer-readable medium |
CN112330037A (en) * | 2020-11-11 | 2021-02-05 | 天津汇商共达科技有限责任公司 | Method and device for predicting inventory proportion of new product and server |
CN113298552A (en) * | 2020-12-02 | 2021-08-24 | 阿里巴巴集团控股有限公司 | Data processing method, server and storage medium |
CN112597745A (en) * | 2020-12-23 | 2021-04-02 | 深圳赛安特技术服务有限公司 | Intelligent chart generation method and device, computer system and readable storage medium |
WO2022199383A1 (en) * | 2021-03-26 | 2022-09-29 | 北京京东拓先科技有限公司 | Article replenishment method and device, and storage medium |
CN113128932A (en) * | 2021-04-16 | 2021-07-16 | 北京京东振世信息技术有限公司 | Warehouse stock processing method and device, storage medium and electronic equipment |
CN113128932B (en) * | 2021-04-16 | 2024-04-16 | 北京京东振世信息技术有限公司 | Warehouse stock processing method and device, storage medium and electronic equipment |
CN113657822A (en) * | 2021-07-20 | 2021-11-16 | 中国邮电器材集团有限公司 | Method and device for distributing commodity quota |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106156880A (en) | A kind of predict the method for inventory allocation ratio, device and electronic equipment | |
Fattahi et al. | Integrated forward/reverse logistics network design under uncertainty with pricing for collection of used products | |
Difrancesco et al. | Optimizing the return window for online fashion retailers with closed-loop refurbishment | |
CN104951468A (en) | Data searching and processing method and system | |
CN110276495A (en) | Commercial articles ordering method, apparatus, computer and storage medium based on artificial intelligence | |
KR101963817B1 (en) | Apparatus and method for generating prediction information based on a keyword search volume | |
CN105096152A (en) | Commodity popularity-based operation execution method and device | |
WO2018142753A1 (en) | Information processing device, method, and program that use deep learning | |
Mittal et al. | Retailer's ordering policy for deteriorating imperfect quality items when demand and price are time-dependent under inflationary conditions and permissible delay in payments | |
Farhat et al. | ERP neural network inventory control | |
Sakib | Restaurant sales prediction using machine learning | |
Banerjee et al. | Implementing E-commerce model for agricultural produce: a research roadmap | |
TW202020781A (en) | Method and device for predicting foreign exchange transaction volume | |
CN107292713A (en) | A kind of rule-based individual character merged with level recommends method | |
Ozmen | Logistics competitiveness of OECD countries using an improved TODIM method | |
Ghanadiof | Customer loyalty and powerful brand in heavy machinery industry | |
Yue et al. | Product life cycle based demand forecasting by using artificial bee colony algorithm optimized two-stage polynomial fitting | |
KR102217886B1 (en) | Exploration System and Method of Optimal Weight of Big Data-based Commodity Investment Recommendation Algorithm Using Artificial Intelligence | |
Gupta et al. | Stock market prediction accuracy analysis using kappa measure | |
Shendryk et al. | Information system for selection the optimal goods supplier | |
Pirani et al. | Analysis and optimization of online sales of products | |
Józefczyk et al. | Heuristic algorithms for the Internet shopping optimization problem with price sensitivity discounts | |
Suprihanti et al. | The impact of clove import policy on clove market and cigarette production in Indonesia | |
KR101409273B1 (en) | Method and apparatus for calculating to measure influential power of users | |
Soonsiripanichkul et al. | Domination dependency analysis of sales marketing based on multi-label classification using label ordering and cycle chain classification |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
TA01 | Transfer of patent application right | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20180423 Address after: Four story 847 mailbox of the capital mansion of Cayman Islands, Cayman Islands, Cayman Applicant after: CAINIAO SMART LOGISTICS HOLDING Ltd. Address before: Cayman Islands Grand Cayman capital building a four storey No. 847 mailbox Applicant before: ALIBABA GROUP HOLDING Ltd. |
|
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161123 |