CN109903064A - Method for Sales Forecast method, apparatus and computer readable storage medium - Google Patents
Method for Sales Forecast method, apparatus and computer readable storage medium Download PDFInfo
- Publication number
- CN109903064A CN109903064A CN201711291842.7A CN201711291842A CN109903064A CN 109903064 A CN109903064 A CN 109903064A CN 201711291842 A CN201711291842 A CN 201711291842A CN 109903064 A CN109903064 A CN 109903064A
- Authority
- CN
- China
- Prior art keywords
- sku
- measured
- sales
- sales forecast
- feature vector
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
Abstract
The invention discloses a kind of Method for Sales Forecast method, apparatus and computer readable storage mediums, are related to data processing field.Method for Sales Forecast method includes: to determine the time attribute of SKU to be measured in the distribution situation of the sales volume information of different time unit according to keeper unit SKU to be measured；The corresponding Method for Sales Forecast model of the time attribute of the feature vector and SKU to be measured that obtain SKU to be measured, wherein the feature vector of SKU to be measured includes the historical data and predicted time information of SKU to be measured；The feature vector of SKU to be measured is input in prediction model, Method for Sales Forecast result is obtained.The accuracy rate of Method for Sales Forecast is improved, cost of labor is reduced in the regularity of distribution of different time it is thus possible to which prediction result is made to meet SKU sales volume to be measured, and then improves the operational efficiency of entire e-commerce platform.
Description
Technical field
The present invention relates to data processing field, in particular to a kind of Method for Sales Forecast method, apparatus and computer-readable storage
Medium.
Background technique
In e-commerce technology, marketing plan carries into execution a plan as sales target, accepts corporate strategic planning upwards
Landing, instruct the operation plan of supplier, storage, logistics downwards, play the role of core hinge.Marketing plan need with
It makes with factors such as market trends and timely adjusting.
Currently, the formulation of marketing plan is mainly formulated by adopting pin personnel by rule of thumb, therefore marketing plan can not be carried out
Fine-grained management, causes that marketing plan accuracy is low, mismatches with actual sales situation, and then affects e-commerce system
The overall operation efficiency of system.
Summary of the invention
One technical problem to be solved by the embodiment of the invention is that: how to promote the accuracy of Method for Sales Forecast.
First aspect according to some embodiments of the invention provides a kind of Method for Sales Forecast method, comprising: according to library to be measured
Storage unit SKU determines the time attribute of SKU to be measured in the distribution situation of the sales volume information of different time unit；It obtains to be measured
The feature vector of SKU and the corresponding Method for Sales Forecast model of the time attribute of SKU to be measured, wherein the feature vector packet of SKU to be measured
Include the historical data and predicted time information of SKU to be measured；The feature vector of SKU to be measured is input in prediction model, is sold
Measure prediction result.
In some embodiments, it is determined to be measured according to SKU to be measured in the distribution situation of the sales volume information of different time unit
The corresponding time attribute of SKU include: be include SKU to be measured multiple SKU generate trend vector, wherein each trend vector packet
Corresponding SKU is included in the sales volume information of each chronomere；The corresponding trend vector of multiple SKU is clustered, according to cluster
As a result SKU generic to be measured determines the time attribute of SKU to be measured in.
In some embodiments, time attribute is season attribute.
In some embodiments, the case where chronomere's quantity that the sales volume information of SKU to be measured is related to is less than preset value
Under, Method for Sales Forecast method further include: be more than default by accounting in multiple item properties of the SKU with same time attribute
The item property of value is determined as the corresponding item property of time attribute；By the item property of SKU to be measured and different time attributes
Corresponding item property is matched；The time attribute of SKU to be measured is determined according to matching result.
In some embodiments, the corresponding multiple pins of the time attribute of the feature vector and SKU to be measured that obtain SKU to be measured
Measure prediction model；The feature vector of SKU to be measured is separately input in multiple Method for Sales Forecast models, according to multiple Method for Sales Forecast moulds
The output result of type obtains Method for Sales Forecast result.
In some embodiments, Method for Sales Forecast knot is obtained according to the weighted sum of the output result of multiple Method for Sales Forecast models
Fruit, weight corresponding to the output result of each Method for Sales Forecast model are determined according to the prediction error of Method for Sales Forecast model.
In some embodiments, multiple Method for Sales Forecast models include the first Method for Sales Forecast model and the second Method for Sales Forecast mould
Type；Method for Sales Forecast result P is obtained using following formula:
Wherein, a is the prediction error of the first Method for Sales Forecast model, and b is the prediction error of the second Method for Sales Forecast model, and g1 is
The output valve of first Method for Sales Forecast model, g2 are the output valve of the second Method for Sales Forecast model.
In some embodiments, Method for Sales Forecast method further include: obtain have same time attribute, for trained
SKU feature vector and corresponding mark value, wherein mark value is the sales volume of corresponding SKU；To be used for trained SKU feature to
Amount is input to in training pattern；According to the gap of the mark value of output result and corresponding SKU feature vector to training pattern
The parameter to training pattern is adjusted, Method for Sales Forecast model is obtained.
In some embodiments, the feature vector of SKU to be measured includes one or more of feature: each period of SKU
History sales volume, SKU each period history be averaged sales volume, SKU in the sales volume of each chronomere of history, SKU in last year
The item property of sales promotion information, SKU that sales volume, the SKU of the same period is related to.
In some embodiments, Method for Sales Forecast method further include: according to the Method for Sales Forecast of multiple SKU to be measured as a result, obtaining
The prediction sales volume ratio of each SKU to be measured；According to the prediction sales volume ratio of plan sales volume total amount and each SKU to be measured, update every
The Method for Sales Forecast result of a SKU to be measured.
The second aspect according to some embodiments of the invention provides a kind of Method for Sales Forecast device, comprising: time attribute is true
Cover half block, for, in the distribution situation of the sales volume information of different time unit, being determined to be measured according to keeper unit SKU to be measured
The time attribute of SKU；Feature vector and model obtain module, for obtain SKU to be measured feature vector and SKU to be measured when
Between the corresponding Method for Sales Forecast model of attribute, wherein when the feature vector of SKU to be measured includes the historical data and prediction of SKU to be measured
Between information；Method for Sales Forecast module obtains Method for Sales Forecast result for the feature vector of SKU to be measured to be input in prediction model.
In some embodiments, time attribute determining module be further used for be include SKU to be measured multiple SKU generate become
Gesture vector, wherein each trend vector includes sales volume information of the corresponding SKU in each chronomere；It is corresponding to multiple SKU
Trend vector is clustered, and the time attribute of SKU to be measured is determined according to SKU generic to be measured in cluster result.
In some embodiments, time attribute is season attribute.
In some embodiments, Method for Sales Forecast device further include: new product time attribute determining module, in SKU to be measured
Chronomere's quantity for being related to of sales volume information be less than preset value in the case where, in the multiple of the SKU with same time attribute
In item property, the item property that accounting is more than preset value is determined as the corresponding item property of time attribute；By SKU's to be measured
Item property item property corresponding from different time attributes is matched；The time of SKU to be measured is determined according to matching result
Attribute.
In some embodiments, feature vector and model obtain module be used to obtain SKU to be measured feature vector and to
Survey the corresponding multiple Method for Sales Forecast models of time attribute of SKU；Method for Sales Forecast module be further used for by the feature of SKU to be measured to
Amount is separately input in multiple Method for Sales Forecast models, obtains Method for Sales Forecast knot according to the output result of multiple Method for Sales Forecast models
Fruit.
In some embodiments, Method for Sales Forecast module is further used for according to the output results of multiple Method for Sales Forecast models
Weighted sum obtains Method for Sales Forecast as a result, weight corresponding to the output result of each Method for Sales Forecast model is according to Method for Sales Forecast model
Prediction error determine.
In some embodiments, multiple Method for Sales Forecast models include the first Method for Sales Forecast model and the second Method for Sales Forecast mould
Type；Method for Sales Forecast module is further used for obtaining Method for Sales Forecast result P using following formula:
Wherein, a is the prediction error of the first Method for Sales Forecast model, and b is the prediction error of the second Method for Sales Forecast model, and g1 is
The output valve of first Method for Sales Forecast model, g2 are the output valve of the second Method for Sales Forecast model.
In some embodiments, Method for Sales Forecast device further includes Method for Sales Forecast model training module, has phase for obtaining
With time attribute, for trained SKU feature vector and corresponding mark value, wherein mark value be corresponding SKU pin
Amount；The SKU feature vector for being used for trained is input to in training pattern；According to the output result to training pattern and accordingly
The gap of the mark value of SKU feature vector adjusts the parameter to training pattern, obtains Method for Sales Forecast model.
In some embodiments, the feature vector of SKU to be measured includes one or more of feature: each period of SKU
History sales volume, SKU each period history be averaged sales volume, SKU in the sales volume of each chronomere of history, SKU in last year
The item property of sales promotion information, SKU that sales volume, the SKU of the same period is related to.
In some embodiments, Method for Sales Forecast device further includes plan sales volume dismantling module, for according to multiple SKU to be measured
Method for Sales Forecast as a result, obtaining the prediction sales volume ratio of each SKU to be measured；According to plan sales volume total amount and each SKU to be measured
It predicts sales volume ratio, updates the Method for Sales Forecast result of each SKU to be measured.
In terms of third according to some embodiments of the invention, a kind of Method for Sales Forecast device is provided, comprising: memory；With
And it is coupled to the processor of the memory, the processor is configured to the instruction based on storage in the memory, holds
Any one aforementioned Method for Sales Forecast method of row.
The 4th aspect according to some embodiments of the invention, provides a kind of computer readable storage medium, stores thereon
There is computer program, which is characterized in that the program realizes any one aforementioned Method for Sales Forecast method when being executed by processor.
Some embodiments in foregoing invention have the following advantages that or the utility model has the advantages that by being belonged to according to the time of SKU to be measured
Property determine Method for Sales Forecast model and automatic Method for Sales Forecast carried out using determining Method for Sales Forecast model, prediction result can be made to accord with
SKU sales volume to be measured is closed in the regularity of distribution of different time, the accuracy rate of Method for Sales Forecast is improved, reduces cost of labor, in turn
Improve the operational efficiency of entire e-commerce platform.
By referring to the drawings to the detailed description of exemplary embodiment of the present invention, other feature of the invention and its
Advantage will become apparent.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention without any creative labor, may be used also for those of ordinary skill in the art
To obtain other drawings based on these drawings.
Fig. 1 is the flow chart according to the Method for Sales Forecast method of some embodiments of the invention.
Fig. 2 is the flow chart that method is determined according to the time attribute of the SKU of some embodiments of the invention.
Fig. 3 is the flow chart that method is determined according to the time attribute of the SKU of other embodiments of the invention.
Fig. 4 is the flow chart according to the method for carrying out Method for Sales Forecast using multiple models of some embodiments of the invention.
Fig. 5 is the flow chart according to the Method for Sales Forecast model training method of some embodiments of the invention.
Fig. 6 is the flow chart according to the Method for Sales Forecast method of other embodiments of the invention.
Fig. 7 is the structure chart according to the Method for Sales Forecast device of some embodiments of the invention.
Fig. 8 is the structure chart according to the Method for Sales Forecast device of other embodiments of the invention.
Fig. 9 is the structure chart according to the Method for Sales Forecast device of yet other embodiments of the invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Below
Description only actually at least one exemplary embodiment be it is illustrative, never as to the present invention and its application or make
Any restrictions.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work premise
Under every other embodiment obtained, shall fall within the protection scope of the present invention.
Unless specifically stated otherwise, positioned opposite, the digital table of the component and step that otherwise illustrate in these embodiments
It is not limited the scope of the invention up to formula and numerical value.
Simultaneously, it should be appreciated that for ease of description, the size of various pieces shown in attached drawing is not according to reality
Proportionate relationship draw.
Technology, method and apparatus known to person of ordinary skill in the relevant may be not discussed in detail, but suitable
In the case of, the technology, method and apparatus should be considered as authorizing part of specification.
It is shown here and discuss all examples in, any occurrence should be construed as merely illustratively, without
It is as limitation.Therefore, the other examples of exemplary embodiment can have different values.
It should also be noted that similar label and letter indicate similar terms in following attached drawing, therefore, once a certain Xiang Yi
It is defined in a attached drawing, then in subsequent attached drawing does not need that it is further discussed.
Fig. 1 is the flow chart according to the Method for Sales Forecast method of some embodiments of the invention.As shown in Figure 1, the embodiment
Method for Sales Forecast method includes step S102~S106.
In step s 102, according to SKU to be measured (Stock Keeping Unit, keeper unit) in different time unit
Sales volume information distribution situation, determine the time attribute of SKU to be measured.
Chronomere can set according to demand, for example, can enable each season is a chronomere, can also enable every
A month or the every two moon are chronomere etc..Chronomere can embody the sales trend of SKU.For example, electric blanket class
SKU it is fast-selling in winter；Mosquito net, the SKU of mosquito-repellent incense class are fast-selling in summer；Steamed crab class SKU is in September, October or so fast sale；Wine
Although class product whole year have a large number of users purchase, around the Spring Festival January, February when due to family dinner party, party
It is more, therefore the sales volume of alcohol product can reach peak value in this period.
In step S104, the corresponding Method for Sales Forecast of time attribute of the feature vector and SKU to be measured of SKU to be measured is obtained
Model, wherein the feature vector of SKU to be measured includes the historical data and predicted time information of SKU to be measured.
The present invention is not all SKU to be measured to be passed through identical model to predict, but press when predicting
Train different Method for Sales Forecast models in advance according to time attribute.That is, in training, using the training number with same time attribute
It is trained according to model.To which more accurate prediction result can be obtained in prediction.
Method for Sales Forecast model for example can for gradient promoted decision tree (Gradient Boosting Decision Tree,
Referred to as: GBDT) model, linear regression model (LRM), length memory network (Long Short-Term Memory, referred to as: LSTM) mould
Type etc..
In the feature vector of SKU to be measured, the historical data of SKU to be measured reflects SKU to be measured in the past period
Sales situation, the data that these historical datas can be the data of SKU itself, be also possible to the affiliated category of the SKU.In some realities
It applies in example, the feature vector of SKU to be measured may include the information that increases by a year-on-year basis of sales volume, such as SKU to be measured is within nearly a period of time
The practical gross turnover (Gross Merchandise Volume, referred to as: GMV) of every month or monthly GMV；In some realities
It applies in example, the feature vector of SKU to be measured may include the sequential growth rate information of sales volume, such as category belonging to SKU to be measured is being gone
The statistical informations such as the GMV of the same period in year or GMV mean value, variance in a period of time of the same period last year.
Predicted time information can be the time corresponding to sales volume to be measured, such as predict the sales volume of winter in this year SKU, then
Winter can be predicted time information；Or the sales volume of prediction SKU in December in this year, then December can be for predicted time information.
Other than above two illustrative mode, predicted time information can take various forms expression, such as date, festivals or holidays
Label etc..If it is date form, either the solar calendar date, being also possible to lunar date.
The feature vector of SKU to be measured can also include the sales promotion information that SKU is related to.Such as the SKU participate in promotion classification,
Promote duration, the SKU quantity of promotion covering, hot-sale products proportion in promotion etc..And promotion of the SKU in competitor's platform
Information can also be added in feature vector.
The feature vector of SKU to be measured can also include the item property, such as specification, size, color etc. of SKU itself.
In step s 106, the feature vector of SKU to be measured is input in prediction model, obtains Method for Sales Forecast result.
Method through the foregoing embodiment can determine Method for Sales Forecast model and be used according to the time attribute of SKU to be measured
Determining Method for Sales Forecast model carries out automatic Method for Sales Forecast, so as to so that prediction result meets SKU sales volume to be measured when different
Between the regularity of distribution, improve the accuracy rate of Method for Sales Forecast, reduce cost of labor, and then improve entire e-commerce platform
Operational efficiency.
In some embodiments, the time attribute of SKU can be determined by the method for cluster.This is described below with reference to Fig. 2
Invent the embodiment of the determination method of the time attribute of SKU.
Fig. 2 is the flow chart that method is determined according to the time attribute of the SKU of some embodiments of the invention.As shown in Fig. 2, should
The time attribute of the SKU of embodiment determines that method includes step S202~S204.
In step S202, trend vector is generated for multiple SKU, wherein each trend vector includes corresponding SKU every
The sales volume information of a chronomere.
The size of chronomere can according to need to set.For example, can be counted as unit of month, if certain
Sales volume of a SKU in 1~December be respectively 1,0.8,0.95,0.98,10.23,15.6,18.9,19.7,10.6,2.5,1,
1.2, unit Wan Yuan, then the trend vector of the SKU be [1,0.8,0.95,0.98,10.23,15.6,18.9,19.7,10.6,
2.5,1,1.2]。
In step S204, the corresponding trend vector of multiple SKU is clustered, according to SKU institute to be measured in cluster result
Belong to the time attribute that classification determines SKU to be measured.
For example, if time attribute be season attribute, classification quantity can be 4, i.e., each classification respectively correspond the spring,
Summer, autumn, four seasons of winter；Quantity of classifying may be 5, i.e., further includes " annual fast-selling " such a class other than the four seasons
Not, sales volume is not changed with time and has the SKU significantly changed to pick out.
Method through the foregoing embodiment, can automatically determine the time attribute of each SKU, thus obtain it is therein to
Survey the time attribute of SKU.The method of above-described embodiment can accurately, objectively identify the trend that the sales volume of SKU changes over time
And rule, so as to accurately identify the time attribute of SKU.
When being clustered, the selling time needs of SKU reache a certain level and can just participate in calculating, i.e., SKU is related to
Chronomere's quantity needs to be greater than preset value.For example, the SKU for having sold 1 year or more has the data of each period in one year,
And for some new categories or new online SKU, sales data is not enough to cover a year and a day, therefore can also use following reality
The method for applying example determines time attribute.The embodiment of the determination method of the time attribute of SKU of the present invention is described below with reference to Fig. 3.
Fig. 3 is the flow chart that method is determined according to the time attribute of the SKU of other embodiments of the invention.As shown in figure 3,
The time attribute of the SKU of the embodiment determines that method includes step S302~S306.
It in step s 302, is more than preset value by accounting in multiple item properties of the SKU with same time attribute
Item property be determined as the corresponding item property of time attribute.
In step s 304, by the item property progress corresponding from different time attributes of the item property of SKU to be measured
Match.
In step S306, the time attribute of SKU to be measured is determined according to matching result.
Such as, it has been determined that the SKU1 with time attribute X have item property A, B, C, SKU2 have item property A, C,
D, SKU3 have item property A, B, E, and SKU2 has item property A, C, F etc..By statistics, the SKU with time attribute X
In, the accounting highest of item property A and C, such as in the SKU with winter attribute, item property " thickening ", " adding suede " commodity
Attribute accounting highest, therefore the SKU with item property A and C is very likely the SKU with time attribute X.
To determine that its time belongs to for the new product SKU that sales data amount cover time length is less than preset value
Property, further improve applicable scene of the invention.After the sale duration of new product SKU reaches preset value, it can also use
Whether Fig. 2 embodiment or similar method verify its time attribute again reasonable.
The embodiment of the present invention Method for Sales Forecast can be determined based on the output result of a model when being predicted as a result,
It can also be determined according to the output result of multiple models.The present invention, which is described, below with reference to Fig. 4 carries out sales volume using multiple models
The embodiment of prediction.
Fig. 4 is the flow chart according to the method for carrying out Method for Sales Forecast using multiple models of some embodiments of the invention.Such as
Shown in Fig. 4, the Method for Sales Forecast model training method of the embodiment includes step S402~S404.
In step S402, the corresponding multiple sales volumes of the time attribute of the feature vector and SKU to be measured that obtain SKU to be measured
Prediction model.For example, SKU to be measured is winter hot-sale products, then available GBDT winter model and LSTM winter model.
In step s 404, the feature vector of SKU to be measured is separately input in multiple Method for Sales Forecast models, according to multiple
The output result of Method for Sales Forecast model obtains Method for Sales Forecast result.
For example, can determine Method for Sales Forecast result according to the weighted results of the output result of multiple Method for Sales Forecast models.It is logical
The method for crossing above-described embodiment can integrate the output of multiple models as a result, improving the accuracy of Method for Sales Forecast.
In some embodiments, Method for Sales Forecast can be obtained according to the weighted sum of the output result of multiple Method for Sales Forecast models
As a result, weight corresponding to the output result of each Method for Sales Forecast model is determined according to the prediction error of Method for Sales Forecast model.Under
Face illustratively provides a kind of preparation method of prediction result.
If being predicted that the two Method for Sales Forecast models are respectively the first Method for Sales Forecast mould based on two Method for Sales Forecast models
Type and the second Method for Sales Forecast model can obtain Method for Sales Forecast result P using formula (1):
Wherein, a is the prediction error of the first Method for Sales Forecast model, and b is the prediction error of the second Method for Sales Forecast model, prediction
Error can export result according to the training stage and the gap of mark value determines；G1 is the output valve of the first Method for Sales Forecast model,
G2 is the output valve of the second Method for Sales Forecast model.It is thus possible to using the error size of model as weight, comprehensively consider multiple moulds
The output of type is as a result, further promote the accuracy of Method for Sales Forecast.
The embodiment of sales volume prediction model training method of the present invention is described below with reference to Fig. 5.
Fig. 5 is the flow chart according to the Method for Sales Forecast model training method of some embodiments of the invention.As shown in figure 5, should
The Method for Sales Forecast model training method of embodiment includes step S502~S506.
In step S502, obtain have same time attribute, for trained SKU feature vector and corresponding mark
Note value, wherein mark value is the sales volume of corresponding SKU.
Feature vector used by training stage is identical as the component of feature vector used by forecast period, here
It repeats no more.Since trained purpose is to obtain the corresponding proprietary model of certain time attribute, training data should also be as having
There is identical time attribute.
In step S504, the SKU feature vector for being used for trained is input to in training pattern.
In step S506, according to the gap of the mark value of output result and corresponding SKU feature vector to training pattern
The parameter to training pattern is adjusted, Method for Sales Forecast model is obtained.
The specific training method of model can refer to training method in the prior art, and which is not described herein again.This field skill
Art personnel, which can according to need, selects model to be trained.Illustratively several models are briefly described below.
GBDT is based on the gradient descent algorithm for promoting (Boosting) integrated approach, and core is that thought is that serial foundation is more
Decision tree, every one tree are all the mistakes for having tree for correcting front, are current to apply more preferable than wide, effect one
Kind machine learning algorithm.
Linear regression is the least square function using referred to as equation of linear regression to one or more independents variable and because becoming
A kind of regression analysis that relationship is modeled between amount.This function is one or more model parameters for being known as regression coefficient
Linear combination.The referred to as simple regression of the case where only one independent variable is called multiple regression greater than independent variable situation.
LSTM is a kind of specific type of recurrent neural network, can learn long-term Dependency Specification.LSTM passes through deliberately
Design is to avoid long-term Dependence Problem.
In the training stage, other than dividing different models according to time attribute, model can also be carried out further
Subdivision.For example, the case where being individually trained to the training data for having promotion resource, there is promotion resource, for example can be headed by
The SKU of page push, the SKU with movable special column, promotion SKU with stronger dynamics etc.；In addition it is also possible to the SKU of new category
Training data carry out individually training etc..The embodiment of sales volume prediction technique of the present invention is described below with reference to Fig. 6.
Fig. 6 is the flow chart according to the Method for Sales Forecast method of other embodiments of the invention.As shown in fig. 6, the embodiment
Method for Sales Forecast method include step S602~S620.
In step S602, according to keeper unit SKU to be measured the sales volume information of different time unit distribution situation,
Determine the time attribute of SKU to be measured.
In step s 604, the feature vector of SKU to be measured is obtained.
In step S606, judge whether SKU to be measured has promotion resource.If so, executing step S608；If it is not,
Execute step S610.
In step S608, using the sales volume of the corresponding promotion model prediction SKU to be measured of the time attribute of SKU to be measured.
Method for Sales Forecast model obtained from there is the training data of promotion resource to be trained according to promotion model.
In step S610, judge whether SKU to be measured is new product.If so, executing step S612；If it is not, executing
Step S614.
In step S612, using the sales volume of the corresponding new product model prediction SKU to be measured of the time attribute of SKU to be measured.
New product model is Method for Sales Forecast model obtained from being trained according to the corresponding training data of new product.
In step S614, the sales volume of SKU to be measured is predicted using the corresponding general models of the time attribute of SKU to be measured.
General models are Method for Sales Forecast mould obtained from being trained according to non-new product, the corresponding training data of non-promotional item
Type.
In step S616, prediction result is loaded into caching, in order to which user is quickly read as needed.
In step S618, marketing plan report is generated according to prediction result, such as can be according to the estimated pin of SKU to be measured
Amount generates the overall planning as unit of adopting pin person, category, brand, department.
It in some embodiments, can be according to the Method for Sales Forecast of multiple SKU to be measured as a result, obtaining the pre- of each SKU to be measured
Survey sales volume ratio；Then further according to the prediction sales volume ratio of plan sales volume total amount and each SKU to be measured, each SKU to be measured is updated
Method for Sales Forecast result.Thus after adopting pin person or part obtains production task, it can be according to the prediction result of SKU
Plan is disassembled, to make prediction result that can more meet actual demand.
In step S620, persistent storage is carried out to data, such as HDFS (Hadoop Distributed can be passed through
File System, distributed file storage system) it saves.
Method through the foregoing embodiment automatically can carry out accurately sales volume for the concrete type of SKU to be measured
Prediction, and data can be carried out to the storage of caching and persistence, it is the links such as the subsequent supply of material, storage, logistics, operation
It provides a great convenience.
The embodiment of sales volume prediction meanss of the present invention is described below with reference to Fig. 7.
Fig. 7 is the structure chart according to the Method for Sales Forecast device of some embodiments of the invention.As shown in fig. 7, the embodiment
Method for Sales Forecast device 70 includes: time attribute determining module 710, is used for according to keeper unit SKU to be measured in different time list
The distribution situation of the sales volume information of position, determines the time attribute of SKU to be measured；Feature vector and model obtain module 720, for obtaining
The corresponding Method for Sales Forecast model of the time attribute of the feature vector and SKU to be measured that take SKU to be measured, wherein the feature of SKU to be measured
Vector includes the historical data and predicted time information of SKU to be measured；Method for Sales Forecast module 730, for by the feature of SKU to be measured to
Amount is input in prediction model, obtains Method for Sales Forecast result.
In some embodiments, time attribute determining module 710 can be further used for being include SKU to be measured multiple
SKU generates trend vector, wherein each trend vector includes sales volume information of the corresponding SKU in each chronomere；To multiple
The corresponding trend vector of SKU is clustered, and the time attribute of SKU to be measured is determined according to SKU generic to be measured in cluster result.
In some embodiments, time attribute is season attribute.
In some embodiments, Method for Sales Forecast device 70 can also include: new product time attribute determining module 740, be used for
In the case where chronomere's quantity that the sales volume information of SKU to be measured is related to is less than preset value, with same time attribute
In multiple item properties of SKU, the item property that accounting is more than preset value is determined as the corresponding item property of time attribute；It will
The item property of SKU to be measured item property corresponding from different time attributes is matched；It is determined according to matching result to be measured
The time attribute of SKU.
In some embodiments, feature vector and model, which obtain module 720, can be used for obtaining the feature vector of SKU to be measured
And the corresponding multiple Method for Sales Forecast models of time attribute of SKU to be measured；Method for Sales Forecast module 730 can be further used for will be to
The feature vector for surveying SKU is separately input in multiple Method for Sales Forecast models, is obtained according to the output result of multiple Method for Sales Forecast models
Obtain Method for Sales Forecast result.
In some embodiments, Method for Sales Forecast module 730 is further used for the output knot according to multiple Method for Sales Forecast models
The weighted sum of fruit obtains Method for Sales Forecast as a result, weight corresponding to the output result of each Method for Sales Forecast model is according to Method for Sales Forecast
The prediction error of model determines.
In some embodiments, multiple Method for Sales Forecast models may include the first Method for Sales Forecast model and the second Method for Sales Forecast
Model；Method for Sales Forecast module 730 can be further used for obtaining Method for Sales Forecast result P using following formula:
Wherein, a is the prediction error of the first Method for Sales Forecast model, and b is the prediction error of the second Method for Sales Forecast model, and g1 is
The output valve of first Method for Sales Forecast model, g2 are the output valve of the second Method for Sales Forecast model.
In some embodiments, Method for Sales Forecast device 70 can also include Method for Sales Forecast model training module 750, for obtaining
Take it is with same time attribute, for trained SKU feature vector and corresponding mark value, wherein mark value is corresponding
The sales volume of SKU；The SKU feature vector for being used for trained is input to in training pattern；According to the output result to training pattern
The parameter to training pattern is adjusted with the gap of the mark value of corresponding SKU feature vector, obtains Method for Sales Forecast model.
In some embodiments, the feature vector of SKU to be measured may include one or more of feature: each of SKU
Be averaged sales volume, SKU of the history sales volume of period, the history of each period of SKU exists in sales volume, the SKU of each chronomere of history
The item property of sales promotion information, SKU that sales volume, the SKU of the same period last year is related to.
In some embodiments, Method for Sales Forecast device 70 can also include that plan sales volume disassembles module 760, for according to more
The Method for Sales Forecast of a SKU to be measured is as a result, obtain the prediction sales volume ratio of each SKU to be measured；According to plan sales volume total amount and each
The prediction sales volume ratio of SKU to be measured updates the Method for Sales Forecast result of each SKU to be measured.
Fig. 8 is the structure chart according to the Method for Sales Forecast device of other embodiments of the invention.As shown in figure 8, the embodiment
Method for Sales Forecast device 800 include: memory 810 and the processor 820 for being coupled to the memory 810, processor 820 is matched
It is set to based on the instruction being stored in memory 810, executes the Method for Sales Forecast method in any one aforementioned embodiment.
Wherein, memory 810 is such as may include system storage, fixed non-volatile memory medium.System storage
Device is for example stored with operating system, application program, Boot loader (Boot Loader) and other programs etc..
Fig. 9 is the structure chart according to the Method for Sales Forecast device of yet other embodiments of the invention.As shown in figure 9, the embodiment
Method for Sales Forecast device 900 include: memory 910 and processor 920, can also include that input/output interface 930, network connect
Mouth 940, memory interface 950 etc..Can for example it lead between these interfaces 930,940,950 and memory 910 and processor 920
Cross the connection of bus 960.Wherein, input/output interface 930 is display, the input-output equipment such as mouse, keyboard, touch screen provide
Connecting interface.Network interface 940 provides connecting interface for various networked devices.Memory interface 950, which is that SD card, USB flash disk etc. are external, to be deposited
It stores up equipment and connecting interface is provided.
The embodiment of the present invention also provides a kind of computer readable storage medium, is stored thereon with computer program, special
Sign is that the program realizes any one aforementioned Method for Sales Forecast method when being executed by processor.
Those skilled in the art should be understood that the embodiment of the present invention can provide as method, system or computer journey
Sequence product.Therefore, complete hardware embodiment, complete software embodiment or combining software and hardware aspects can be used in the present invention
The form of embodiment.Moreover, it wherein includes the calculating of computer usable program code that the present invention, which can be used in one or more,
Machine can use the meter implemented in non-transient storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of calculation machine program product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It is interpreted as to be realized by computer program instructions each in flowchart and/or the block diagram
The combination of process and/or box in process and/or box and flowchart and/or the block diagram.It can provide these computer journeys
Sequence instruct to general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices processor with
A machine is generated, so that the instruction generation executed by computer or the processor of other programmable data processing devices is used for
Realize the dress for the function of specifying in one or more flows of the flowchart and/or one or more blocks of the block diagram
It sets.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and
Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (20)
1. a kind of Method for Sales Forecast method, comprising:
According to keeper unit SKU to be measured in the distribution situation of the sales volume information of different time unit, the time of SKU to be measured is determined
Attribute；
Obtain the corresponding Method for Sales Forecast model of time attribute of the feature vector and the SKU to be measured of SKU to be measured, wherein institute
The feature vector for stating SKU to be measured includes the historical data and predicted time information of SKU to be measured；
The feature vector of the SKU to be measured is input in the prediction model, Method for Sales Forecast result is obtained.
2. Method for Sales Forecast method according to claim 1, wherein it is described according to SKU to be measured different time unit pin
The distribution situation for measuring information, determines that the corresponding time attribute of SKU to be measured includes:
It is the multiple SKU generation trend vector for including SKU to be measured, wherein each trend vector includes corresponding SKU when each
Between unit sales volume information；
The corresponding trend vector of the multiple SKU is clustered, is determined according to SKU generic to be measured in cluster result to be measured
The time attribute of SKU.
3. Method for Sales Forecast method according to claim 1 or 2, wherein the time attribute is season attribute.
4. Method for Sales Forecast method according to claim 1, wherein in the chronomere that the sales volume information of SKU to be measured is related to
In the case that quantity is less than preset value, the Method for Sales Forecast method further include:
In multiple item properties of the SKU with same time attribute, the item property that accounting is more than preset value is determined as
The corresponding item property of the time attribute；
The item property of SKU to be measured item property corresponding from different time attributes is matched；
The time attribute of SKU to be measured is determined according to matching result.
5. Method for Sales Forecast method according to claim 1, wherein
The corresponding multiple Method for Sales Forecast models of the time attribute of the feature vector and the SKU to be measured that obtain SKU to be measured；
The feature vector of the SKU to be measured is separately input in multiple Method for Sales Forecast models, according to multiple Method for Sales Forecast models
Output result obtain Method for Sales Forecast result.
6. Method for Sales Forecast method according to claim 5, wherein according to adding for the output result of multiple Method for Sales Forecast models
It weighs and obtains Method for Sales Forecast as a result, weight corresponding to the output result of each Method for Sales Forecast model is according to Method for Sales Forecast model
Predict that error determines.
7. Method for Sales Forecast method according to claim 1, further includes:
Obtain have same time attribute, for trained SKU feature vector and corresponding mark value, wherein the mark
Note value is the sales volume of corresponding SKU；
It is used for trained SKU feature vector by described and is input to in training pattern；
Gap adjustment according to the output result to training pattern and the mark value of corresponding SKU feature vector is described wait train
The parameter of model obtains Method for Sales Forecast model.
8. Method for Sales Forecast method according to claim 1, wherein the feature vector of the SKU to be measured includes following one kind
Or various features:
The history sales volume of each period of SKU, the history of each period of SKU are averaged sales volume, SKU in each chronomere of history
Sales volume, SKU in the same period last year of sales volume, SKU be related to sales promotion information, SKU item property.
9. Method for Sales Forecast method according to claim 1, further includes:
According to the Method for Sales Forecast of multiple SKU to be measured as a result, obtaining the prediction sales volume ratio of each SKU to be measured；
According to the prediction sales volume ratio of plan sales volume total amount and each SKU to be measured, the Method for Sales Forecast knot of each SKU to be measured is updated
Fruit.
10. a kind of Method for Sales Forecast device, comprising:
Time attribute determining module, for according to keeper unit SKU to be measured the sales volume information of different time unit distribution
Situation determines the time attribute of SKU to be measured；
Feature vector and model obtain module, for obtaining the feature vector of SKU to be measured and the time attribute of the SKU to be measured
Corresponding Method for Sales Forecast model, wherein the feature vector of the SKU to be measured includes the historical data and predicted time of SKU to be measured
Information；
Method for Sales Forecast module obtains Method for Sales Forecast for the feature vector of the SKU to be measured to be input in the prediction model
As a result.
11. Method for Sales Forecast device according to claim 10, wherein the time attribute determining module be further used for for
Multiple SKU including SKU to be measured generate trend vector, wherein each trend vector includes corresponding SKU in each chronomere
Sales volume information；The corresponding trend vector of the multiple SKU is clustered, according to SKU generic to be measured in cluster result
Determine the time attribute of SKU to be measured.
12. Method for Sales Forecast device described in 0 or 11 according to claim 1, wherein the time attribute is season attribute.
13. Method for Sales Forecast device according to claim 10, further includes:
New product time attribute determining module, the chronomere's quantity being related to for the sales volume information in SKU to be measured are less than preset value
In the case where, in multiple item properties of the SKU with same time attribute, the item property for being more than preset value for accounting is true
It is set to the corresponding item property of the time attribute；By the item property of SKU to be measured commodity corresponding from different time attributes
Attribute is matched；The time attribute of SKU to be measured is determined according to matching result.
14. Method for Sales Forecast device according to claim 10, wherein
Described eigenvector and model obtain module and are used to obtain the feature vector of SKU to be measured and the time of the SKU to be measured
The corresponding multiple Method for Sales Forecast models of attribute；
The Method for Sales Forecast module is further used for the feature vector of the SKU to be measured being separately input to multiple Method for Sales Forecast moulds
In type, Method for Sales Forecast result is obtained according to the output result of multiple Method for Sales Forecast models.
15. Method for Sales Forecast device according to claim 14, wherein the Method for Sales Forecast module is further used for wherein,
According to the weighted sum of the output result of multiple Method for Sales Forecast models acquisition Method for Sales Forecast as a result, the output of each Method for Sales Forecast model
As a result corresponding weight is determined according to the prediction error of Method for Sales Forecast model.
16. Method for Sales Forecast device according to claim 10 further includes Method for Sales Forecast model training module, for obtaining tool
Have same time attribute, for trained SKU feature vector and corresponding mark value, wherein the mark value is corresponding
The sales volume of SKU；It is used for trained SKU feature vector by described and is input to in training pattern；According to described to training pattern
The gap adjustment parameter to training pattern for exporting the mark value of result and corresponding SKU feature vector, obtains Method for Sales Forecast
Model.
17. Method for Sales Forecast device according to claim 10, wherein the feature vector of the SKU to be measured includes with next
Kind or various features:
The history sales volume of each period of SKU, the history of each period of SKU are averaged sales volume, SKU in each chronomere of history
Sales volume, SKU in the same period last year of sales volume, SKU be related to sales promotion information, SKU item property.
Further include plan sales volume dismantling module 18. Method for Sales Forecast device according to claim 10, for according to it is multiple to
The Method for Sales Forecast of SKU is surveyed as a result, obtaining the prediction sales volume ratio of each SKU to be measured；According to plan sales volume total amount and each to be measured
The prediction sales volume ratio of SKU updates the Method for Sales Forecast result of each SKU to be measured.
19. a kind of Method for Sales Forecast device, in which:
Memory；And
It is coupled to the processor of the memory, the processor is configured to the instruction based on storage in the memory,
Execute such as Method for Sales Forecast method according to any one of claims 1 to 9.
20. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
Method for Sales Forecast method according to any one of claims 1 to 9 is realized when execution.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711291842.7A CN109903064A (en) | 2017-12-08 | 2017-12-08 | Method for Sales Forecast method, apparatus and computer readable storage medium |
| PCT/CN2018/115652 WO2019109790A1 (en) | 2017-12-08 | 2018-11-15 | Sales volume prediction method and device, and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711291842.7A CN109903064A (en) | 2017-12-08 | 2017-12-08 | Method for Sales Forecast method, apparatus and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109903064A true CN109903064A (en) | 2019-06-18 |
Family
ID=66751301
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711291842.7A Pending CN109903064A (en) | 2017-12-08 | 2017-12-08 | Method for Sales Forecast method, apparatus and computer readable storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109903064A (en) |
| WO (1) | WO2019109790A1 (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111160968A (en) * | 2019-12-27 | 2020-05-15 | 清华大学 | SKU-level commodity sales prediction method and device |
| CN111652655A (en) * | 2020-06-10 | 2020-09-11 | 创新奇智(上海)科技有限公司 | Commodity sales prediction method and device, electronic equipment and readable storage medium |
| CN112906925A (en) * | 2019-11-19 | 2021-06-04 | 北京沃东天骏信息技术有限公司 | Goods replenishment method, device and computer-readable storage medium |
| WO2021139335A1 (en) * | 2020-07-28 | 2021-07-15 | 平安科技（深圳）有限公司 | Method and apparatus for predicting sales data of physical machine, and computer device and storage medium |
| CN113487359A (en) * | 2021-07-12 | 2021-10-08 | 润联软件系统(深圳)有限公司 | Multi-modal feature-based commodity sales prediction method and device and related equipment |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113283936A (en) * | 2021-05-28 | 2021-08-20 | 深圳千岸科技股份有限公司 | Sales forecasting method, sales forecasting device and electronic equipment |
| CN113379125B (en) * | 2021-06-11 | 2022-05-13 | 武汉大学 | Logistics storage sales prediction method based on TCN and LightGBM combined model |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1423215A (en) * | 2001-11-27 | 2003-06-11 | 株式会社世界 | Market forecasting device and method thereof |
| CN102968670A (en) * | 2012-10-23 | 2013-03-13 | 北京京东世纪贸易有限公司 | Method and device for predicting data |
| US20140200992A1 (en) * | 2013-01-14 | 2014-07-17 | Oracle International Corporation | Retail product lagged promotional effect prediction system |
| CN106408341A (en) * | 2016-09-21 | 2017-02-15 | 北京小米移动软件有限公司 | Goods sales volume prediction method and device, and electronic equipment |
| CN106570573A (en) * | 2015-10-13 | 2017-04-19 | 阿里巴巴集团控股有限公司 | Parcel attribute information prediction method and device |
| CN106971249A (en) * | 2017-05-05 | 2017-07-21 | 北京挖玖电子商务有限公司 | A kind of Method for Sales Forecast and replenishing method |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104700152B (en) * | 2014-10-22 | 2018-08-10 | 浙江中烟工业有限责任公司 | A kind of tobacco Method for Sales Forecast method of fusion season sales information and search behavior information |
| US20160117702A1 (en) * | 2014-10-24 | 2016-04-28 | Vedavyas Chigurupati | Trend-based clusters of time-dependent data |
| CN106951514A (en) * | 2017-03-17 | 2017-07-14 | 合肥工业大学 | A kind of automobile Method for Sales Forecast method for considering brand emotion |
-
2017
- 2017-12-08 CN CN201711291842.7A patent/CN109903064A/en active Pending
-
2018
- 2018-11-15 WO PCT/CN2018/115652 patent/WO2019109790A1/en active Application Filing
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1423215A (en) * | 2001-11-27 | 2003-06-11 | 株式会社世界 | Market forecasting device and method thereof |
| CN102968670A (en) * | 2012-10-23 | 2013-03-13 | 北京京东世纪贸易有限公司 | Method and device for predicting data |
| US20140200992A1 (en) * | 2013-01-14 | 2014-07-17 | Oracle International Corporation | Retail product lagged promotional effect prediction system |
| CN106570573A (en) * | 2015-10-13 | 2017-04-19 | 阿里巴巴集团控股有限公司 | Parcel attribute information prediction method and device |
| CN106408341A (en) * | 2016-09-21 | 2017-02-15 | 北京小米移动软件有限公司 | Goods sales volume prediction method and device, and electronic equipment |
| CN106971249A (en) * | 2017-05-05 | 2017-07-21 | 北京挖玖电子商务有限公司 | A kind of Method for Sales Forecast and replenishing method |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112906925A (en) * | 2019-11-19 | 2021-06-04 | 北京沃东天骏信息技术有限公司 | Goods replenishment method, device and computer-readable storage medium |
| CN111160968A (en) * | 2019-12-27 | 2020-05-15 | 清华大学 | SKU-level commodity sales prediction method and device |
| CN111652655A (en) * | 2020-06-10 | 2020-09-11 | 创新奇智(上海)科技有限公司 | Commodity sales prediction method and device, electronic equipment and readable storage medium |
| WO2021139335A1 (en) * | 2020-07-28 | 2021-07-15 | 平安科技（深圳）有限公司 | Method and apparatus for predicting sales data of physical machine, and computer device and storage medium |
| CN113487359A (en) * | 2021-07-12 | 2021-10-08 | 润联软件系统(深圳)有限公司 | Multi-modal feature-based commodity sales prediction method and device and related equipment |
| CN113487359B (en) * | 2021-07-12 | 2024-03-22 | 华润数字科技有限公司 | Commodity sales predicting method and device based on multi-mode characteristics and related equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019109790A1 (en) | 2019-06-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109903064A (en) | Method for Sales Forecast method, apparatus and computer readable storage medium | |
| US11055731B2 (en) | Parallel processing historical data | |
| US20050192930A1 (en) | System and method of real estate data analysis and display to support business management | |
| WO2019055065A1 (en) | Evaluating models that rely on aggregate historical data | |
| JPWO2018207259A1 (en) | Information processing system, information processing apparatus, prediction model extraction method, and prediction model extraction program | |
| CN109858934A (en) | Pricing method, device and computer readable storage medium | |
| CN108629632A (en) | Predict the method, apparatus and computer readable storage medium of user's income | |
| CN107918813A (en) | Trend prediction analysis method, equipment and storage medium | |
| Vlckova et al. | Role of demand planning in business process management | |
| Benth et al. | A space-time random field model for electricity forward prices | |
| CN104252344A (en) | Method utilizing template for batched generation of contracts | |
| CN110163683A (en) | Value user's key index determines method, advertisement placement method and device | |
| CN107085774A (en) | A kind of sale of electricity service platform method of evaluating performance and device | |
| CN109886737A (en) | Needing forecasting method, device, electronic equipment and readable storage medium storing program for executing | |
| CN107464152A (en) | Data processing method, client and service end | |
| Zhao et al. | A state‐feedback approach to inventory control: Analytical and empirical studies | |
| JP2016033704A (en) | Income summary prediction device and income summary prediction program | |
| Liebert | Airport benchmarking: an efficiency analysis of European airports from an economic and managerial perspective | |
| WO2023020255A1 (en) | Data processing method and apparatus, device, and storage medium | |
| CN113435541B (en) | Method and device for planning product classes, storage medium and computer equipment | |
| CN109636444A (en) | A kind of cigarette source of goods automatic measurement & calculation and jettison system | |
| Soegoto et al. | Implementation of E-Budgeting Information System on Budget Management PT. Industri Telekomunikasi Indonesia, Indonesia | |
| CN108665097A (en) | A kind of freight demand simulating and predicting method, device and storage medium | |
| US20150006342A1 (en) | Generating a Simulated Invoice | |
| Sillanpää et al. | Critical attributes on supply chain strategy implementation: case study in Europe and Asia |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190618 |
|
| RJ01 | Rejection of invention patent application after publication |