CN109829818A - Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing - Google Patents
Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing Download PDFInfo
- Publication number
- CN109829818A CN109829818A CN201910108630.3A CN201910108630A CN109829818A CN 109829818 A CN109829818 A CN 109829818A CN 201910108630 A CN201910108630 A CN 201910108630A CN 109829818 A CN109829818 A CN 109829818A
- Authority
- CN
- China
- Prior art keywords
- cash
- sales counter
- allocated
- prediction
- demand
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of cash demand amount prediction techniques, device, electronic equipment and readable storage medium storing program for executing, obtain testing data, the testing data includes: the cash time to be allocated, the sales counter geographic location information of cash to be allocated, withdrawal affecting parameters, the history cash demand change information of the sales counter, by the cash demand prediction model of testing data input prebuild, the corresponding prediction cash to be allocated of the sales counter is determined by the cash demand prediction model, wherein, cash demand prediction model has the ability for the actually required minimum cash that the prediction cash to be allocated of the sales counter is tended to sales counter.Actually required minimum cash refers under the premise of can satisfy user's withdrawal demand, so that the cash surplus of sales counter is minimum.Not only can be excessive to avoid cash is stored in sales counter to realize, but also can satisfy the purpose of user's withdrawal demand.
Description
Technical field
The present invention relates to financial technology field, more specifically, it relates to a kind of cash demand amount prediction technique, device,
Electronic equipment and readable storage medium storing program for executing.
Background technique
Cash is stored sometimes in sales counter in financial industry, for example, depositing in automatic sales counter (i.e. ATM machine) or artificial sales counter
Put cash;If very few in the cash of sales counter storage, the withdrawal demand of user cannot be met in time;If in the cash of sales counter storage
Excessively, then the cash for being stored in sales counter, which cannot be used for the modes such as investment, loan, creates more values.
To sum up, cash demand amount needed for how determining sales counter is the technical problem of those skilled in the art.
Summary of the invention
In view of this, the present invention provides a kind of cash demand amount prediction technique, device, electronic equipment and readable storages
Medium, the invention provides the following technical scheme:
A kind of cash demand amount prediction technique, comprising:
Obtain testing data, the testing data include: cash time to be allocated, cash to be allocated sales counter where it is geographical
Location information, withdrawal affecting parameters, the sales counter history cash demand change information;
It is true by the cash demand prediction model by the cash demand prediction model of testing data input prebuild
The fixed sales counter is corresponding to predict cash to be allocated;
Wherein, the cash demand prediction model, which has, tends to the sales counter for the prediction cash to be allocated of the sales counter
The ability of actually required minimum cash.
A kind of cash demand amount prediction meanss, comprising:
First obtains module, and for obtaining testing data, the testing data includes: cash time to be allocated, to be allocated
The sales counter geographic location information of cash, withdrawal affecting parameters, the sales counter history cash demand change information;
Determining module, for passing through the cash for the cash demand prediction model of testing data input prebuild
Demand Forecast Model determines the corresponding prediction cash to be allocated of the sales counter;
Wherein, the cash demand prediction model, which has, tends to the sales counter for the prediction cash to be allocated of the sales counter
The ability of actually required minimum cash.
A kind of electronic equipment, comprising:
Memory, for storing program;
Processor, for executing described program, described program is specifically used for:
Obtain testing data, the testing data include: cash time to be allocated, cash to be allocated sales counter where it is geographical
Location information, withdrawal affecting parameters, the sales counter history cash demand change information;
It is true by the cash demand prediction model by the cash demand prediction model of testing data input prebuild
The fixed sales counter is corresponding to predict cash to be allocated;
Wherein, the cash demand prediction model, which has, tends to the sales counter for the prediction cash to be allocated of the sales counter
The ability of actually required minimum cash.
A kind of readable storage medium storing program for executing, is stored thereon with computer program, which is characterized in that the computer program is processed
When device executes, each step that cash demand amount prediction technique as described above includes is realized.
It can be seen via above technical scheme that compared with prior art, the invention discloses a kind of predictions of cash demand amount
Method, obtains testing data, and testing data includes: the sales counter geographic location letter of cash time to be allocated, cash to be allocated
The history cash demand change information of breath, withdrawal affecting parameters, sales counter;The cash demand of testing data input prebuild is pre-
Model is surveyed, the corresponding prediction cash to be allocated of sales counter is determined by cash demand prediction model;Wherein, cash demand prediction model
Ability with the actually required minimum cash that the prediction cash to be allocated of the sales counter is tended to sales counter.It is actually required minimum existing
Gold refers under the premise of can satisfy user's withdrawal demand, so that the cash surplus of sales counter is minimum.It both can be with to realize
It avoids storing cash in sales counter excessive, and can satisfy the purpose of user's withdrawal demand.
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
The embodiment of invention for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is a kind of implementation flow chart of cash demand amount prediction technique disclosed by the embodiments of the present invention;
It is bent by cash demand variation as unit of day within two weeks that Fig. 2 illustrates certain bank provided by the embodiments of the present application
Line chart;
Fig. 3 illustrates neural network training process schematic diagram provided by the embodiments of the present application;
Fig. 4 is a kind of cash demand amount prediction meanss structural schematic diagram disclosed by the embodiments of the present invention;
Fig. 5 illustrates a kind of structure chart of implementation of electronic equipment provided by the embodiments of the present application.
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.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Financial industry needs to store cash in sales counter, such as in artificial sales counter storage cash or in automatic sales counter (i.e. ATM
Machine) cash is stored in paper money case, but artificially carried out in advance since the prediction of previous sales counter cash to be allocated is all based on artificial experience
It surveys;Cause to predict cash to be allocated that there is a certain error with actual required minimum cash, predict it is not accurate enough, thus
Cause the problem that the storage of sales counter cash is excessive or storage is insufficient.
The actually required minimum cash that the embodiment of the present application refers to refers under the premise of meeting user's withdrawal demand, cabinet
Minimum cash amount needed for platform.
This programme proposes a kind of cash demand amount prediction technique, can determine that sales counter is corresponding by cash demand prediction model
Prediction cash to be allocated, since cash demand prediction model has the reality that the prediction cash to be allocated of sales counter is tended to sales counter
The ability of required minimum cash, so that realizing not only can be excessive to avoid cash is stored in sales counter, but also can satisfy user's withdrawal need to
The purpose asked.
Cash demand amount prediction technique provided by the embodiments of the present application can be applied to electronic equipment, which can be with
For desktop computer or the electronic equipments such as mobile terminal (such as smart phone) or ipad.
Optionally, which can also be a server, or, a server cluster, or, a cloud computing clothes
Business center.
Processing unit provided by the embodiments of the present application can be the client operated in electronic equipment, which can be with
It is application client, is also possible to webpage client.
The embodiment of the present application, which provides cash demand amount prediction technique, can be applied to following application scenarios:
In the first application scenarios, cash demand amount prediction technique provided by the embodiments of the present application can be applied to branch and manage
Cash demand amount forecasting system in reason system, which can be set in each bank outlets, only right
The sales counter of this site carries out the prediction of cash demand amount.
In the second application scenarios, cash demand amount prediction technique provided by the embodiments of the present application can be applied to head office's pipe
Cash demand amount forecasting system in reason system, the cash demand amount forecasting system may include that several terminal devices form
Device clusters, such as server cluster, including a central server and several branch office service devices.Central server
Bank Headquarters can be located at, branch office service device is located at branch, bank, and central server is connected with each branch office service device, passes through branch
Server obtains the testing data of corresponding sales counter, which is inputted cash demand prediction model, obtained by central server
The corresponding prediction cash to be allocated of the testing data, is sent to corresponding branch office service device for prediction cash to be allocated.
Attached drawing 1 is please referred to, attached drawing 1 is a kind of implementation process of cash demand amount prediction technique disclosed by the embodiments of the present invention
Figure, this method include: in detail
Step S101, testing data is obtained, the testing data includes: the cabinet of cash time to be allocated, cash to be allocated
Platform geographic location information, withdrawal affecting parameters, the sales counter history cash demand change information.
The testing data referred in following introduction step S101:
The cash time to be allocated, that is, need to predict the time of cash to be allocated;Wherein it is possible to according to cash to be allocated is predicted
Predetermined period (optional, predetermined period can be 1 day, the .. or, 2 days or 3 days ...) reasonably select the cash time to be allocated,
For example, current time is on December 9th, 2018, predetermined period of the cash to be allocated of certain bank is one day, then needs on the day of
The cash to be allocated in next day (on December 10th, 2018) is predicted, then the cash time to be allocated is on December 10th, 2018.If
Predetermined period of the cash demand amount of bank is one week, then in current all (on December 9,3 days to 2018 December in 2018), and
Current time is on December 9th, 2018 need to predict next all (on December 16,10 days to 2018 December in 2018) to be allocated
Cash, then the cash time to be allocated can be a time range, for example, being December 16 10 days to 2018 December in 2018
Day.
Optionally, the cash time to be allocated includes: exact date and the characterization date is working day or weekend or festivals or holidays
Or the parameter of electric business consumption day;Alternatively, the cash time to be allocated includes: exact date.
Optionally, the sales counter geographic location information of cash to be allocated can be the latitude and longitude information where the sales counter,
Alternatively, the street information where the sales counter of cash to be allocated.
It is understood that the bustling degree of diverse geographic location is different, for example, the prosperity in Shanghai and Hebei province Baoding
Degree is different, for another example Hebei province, southern area, Baoding is different with the bustling degree of Beishi District, optionally, cash to be allocated
The bustling degree of bank outlets where sales counter geographic location information can characterize the sales counter.
Optionally, the sales counter of diverse geographic location information can be numbered in advance, this is inquired by number information
The bustling degree of bank outlets where sales counter.
It is understood that bustling degree is different, the degree of consumption of user is different.
Optionally, withdrawal affecting parameters include: the benchmark interest rate and/or inhabitant's consumption level of the sales counter of cash to be allocated
The Macroscopic Factors such as index, it is clear that the floating of benchmark interest rate and the height of inhabitant's consumption level can influence the access amount of money of client
Volume.
The history cash demand change information of sales counter, the i.e. running parameter of cash demand in time domain scale, can define
For the time domain factor, optionally, history cash demand change information includes: in the previous preset time period of cash time to be allocated
Subject to history cash demand change information, the peak of the actually required minimum cash of sales counter in the preset time period, minimum is selected
It is one or more in value, mean value, variance, root-mean-square value, flexure and kurtosis.
The previous preset time period that the distribution cash time is treated in citing below is illustrated.
Assuming that current time is on December 9th, 2018, the cash time to be allocated is on December 10th, 2018, then previous pre-
If the period refers to the preset time period in the time range for belonging on December 9th, 1, preset time period pair
The time range answered can for 1 day or 2 days or 3 days or 4 days or ....
Optionally, time interval can be divided with chronomere within a preset period of time, for example, with 1 day or 2 days or 3 days
Or 4 days, or ... be chronomere, then within a preset period of time the peak of the actually required minimum cash of sales counter, minimum,
Value, variance, root-mean-square value, flexure and kurtosis refer to the corresponding peak of each time interval, minimum within a preset period of time
Value, mean value, variance, root-mean-square value, flexure and kurtosis.
For example, chronomere is 2 days, it is assumed that preset time period is [on December 2nd, 2018, on December 9th, 2018], that
, obtained each time interval are as follows: [on December 2nd, 2018, on December 3rd, 2018], [on December 4th, 2018, in December, 2018
5 days], [on December 6th, 2018, on December 7th, 2018], [on December 8th, 2018, on December 9th, 2018].Obtain each time zone
The actually required minimum cash of the corresponding sales counter in domain, by taking time interval [on December 2nd, 2018, on December 3rd, 2018] as an example,
It is assumed that the actually required minimum cash of sales counter on December 2nd, 2018 be 10w, on December 3rd, 2018 the actually required minimum cash of sales counter
For 20w, then the corresponding actually required minimum cash of sales counter of time interval [on December 2nd, 2018, on December 3rd, 2018]=
10w+20w=30w, other times section is similar, here without repeating.
Based on the actually required minimum cash of the corresponding sales counter of each time interval, the highest in preset time period is obtained
Value, minimum, mean value, variance, root-mean-square value, flexure and kurtosis.
Optionally, the maximum value in preset time period refers to that the corresponding sales counter of each time interval is actually required minimum existing
The maximum value of gold;Minimum value in preset time period refers to the actually required minimum cash of the corresponding sales counter of each time interval
Minimum value.
Optionally, mean value is average for the corresponding actually required minimum cash of sales counter of time interval each in preset time period
ValueCalculation formula are as follows:
Wherein, n refers to the total number of each time interval in preset time period;I refers to i-th of time interval;xiRefer to i-th
The corresponding actually required minimum cash of sales counter of a time interval.
It optionally, can be according to mean valueVariance α, root-mean-square value is calculatedFlexure β and kurtosis λ, calculation formula
It is as follows:
Assuming that current time is on December 15th, 2018, preset time period is [on December 1st, 2018, in December, 2018
14], chronomere is 1 day, then time interval has 14, specifically as shown in Fig. 2, Fig. 2 illustrates certain bank when this is preset
Between section the actually required minimum cash change curve of sales counter (Fig. 2 corresponds to data shown in table 1), abscissa is day in Fig. 2
Phase, ordinate is the actually required minimum cash of sales counter, in conjunction with Fig. 2 and table 1, available conclusion: in the preset time period most
High level is the 260000 of December 14, and minimum is the 130000 of December 09, is computed, and mean value is 17.36 ten thousand, variance 15.37
Ten thousand, root-mean-square value is 17.79 ten thousand, and flexure is 43.84 ten thousand, and kurtosis is 613.09 ten thousand, wherein the actually required minimum cash of sales counter is
It rounds up as unit of ten thousand, and each value that history cash demand change information includes retains two-decimal as unit of ten thousand.
Table 1
Optionally, testing data can also include: the cash that world economic situation parameter and sales counter are at best able to distribution
One of number is a variety of.
Wherein, world economic situation parameter includes: world's exchange rate, stock market, fund, political setting, one in crude oil price
Kind is a variety of.
Optionally, the actually required minimum cash of sales counter should be less than or equal to the cash number that sales counter is at best able to distribution.
Step S102, by the cash demand prediction model of testing data input prebuild, pass through the cash demand
Prediction model determines the corresponding prediction cash to be allocated of the sales counter;Wherein, have will be described for the cash demand prediction model
The prediction cash to be allocated of sales counter tends to the ability of the actually required minimum cash of the sales counter.
The site referred in the embodiment of the present application can be a bank outlets, alternatively, having for financial industry is deposited
The site of withdrawal demand, is illustrated by taking bank outlets as an example below.
One bank outlets may be provided with one or more artificial sales counters, and/or, one or more ATM machine.
Optionally, cash demand prediction model is the prediction that cash to be allocated is carried out for any one ATM machine, i.e. cash needs
The prediction cash to be allocated for asking prediction model to export refers to the corresponding prediction cash to be allocated of the ATM machine;Alternatively, cash demand is pre-
Surveying model is the prediction that cash to be allocated is carried out for all ATM machine of a bank outlets, i.e. cash demand prediction model is defeated
Cash summation needed for the cash to be allocated of prediction out refers to all ATM machine of the bank outlets.
Optionally, cash demand prediction model is the prediction that cash to be allocated is carried out for any one artificial sales counter, i.e., existing
The prediction cash to be allocated of golden Demand Forecast Model output refers to that the artificial sales counter is corresponding and predicts cash to be allocated;Alternatively, existing
Golden Demand Forecast Model is that the prediction of cash to be allocated, i.e. cash demand are carried out for all artificial sales counters of a bank outlets
Cash summation needed for the prediction cash to be allocated of prediction model output refers to all artificial sales counters of the bank outlets.
Optionally, in step s 102, the sales counter type that output result is directed to is different, or, the sales counter number being directed to is different,
Cash demand prediction model is different, different for the testing data of different cash demand prediction models input.
For example, if the sales counter referred in step S101 is an ATM machine, then the cash demand prediction model is to be based on being somebody's turn to do
The corresponding sample data training of ATM machine obtains;The prediction cash to be allocated of cash demand prediction model output is for the ATM
Machine, optionally, it is not particularly suited for other ATM machine.
Optionally, by withdraw the money for, if a bank outlets are provided with multiple ATM machine, user when withdrawing the money from ATM machine,
It may be to withdraw the money at random from any one ATM machine;Alternatively, if an ATM machine without remaining sum, then user can be from another
ATM machine is withdrawn the money, so the corresponding sample data of an ATM machine can indicate each ATM that the site includes to a certain extent
The corresponding sample data of machine;In this case, the cash demand prediction model obtained with the corresponding sample data training of an ATM
It can be adapted for any ATM machine in the bank outlets.
For another example if the sales counter referred in step S101 is multiple ATM machine, by taking two ATM machine as an example, then the cash needs
Seeking prediction model is obtained based on the corresponding sample data training of two ATM machine;The prediction of cash demand prediction model output
Cash to be allocated is that two ATM machine distinguish the sum of actually required minimum cash.For shown in table 1 and Fig. 2, to two ATM machine
(respectively ATM1 and ATM2) corresponding sample data is illustrated, it is assumed that on December 1st, 2018, ATM1 was actually required minimum
Cash is 10w, and the actually required minimum cash of ATM2 is 5w, then two ATM machine corresponding 1 day December in 2018 is actually required
Minimum cash is 15W, is the sum of corresponding actually required minimum cash of two ATM machine.Obtaining history cash demand
When change information, it is also based on the sum of corresponding actually required minimum cash of two each time intervals of ATM machine and obtains.
If the sales counter referred in step S101 is an artificial sales counter or multiple artificial sales counters, above-mentioned be directed to may refer to
The description of ATM machine, which is not described herein again.
Below to " cash demand prediction model has the reality that the prediction cash to be allocated of the sales counter is tended to the sales counter
The reason of ability of minimum cash needed for border ", is illustrated.
Optionally, which obtained using multiple sample datas training neural network.
Wherein cash demand prediction model is the cash demand that training neural network obtains using sample data as training data
Prediction model;It can choose various forms of neural networks in advance, training neural network generally comprises three elements: structure, algorithm
And weight, neural network can then determine the structure and algorithm in the three elements of the neural network, utilize sample once selecting
The process of the selected neural network of data training is the process being adjusted to the weight in neural network, and detailed process can
To include: that sample data is inputted the neural network, prediction cash to be allocated which is exported according to neural network with
The difference of the corresponding actually required minimum cash of the sample data adjusts the weight of neural network.
Just have by the cash demand prediction model that iteration update training obtains as a result, the prediction of sales counter is to be allocated existing
Gold tends to the ability of the actually required minimum cash of sales counter, so mould is predicted in the cash demand of testing data input prebuild
Type determines that the corresponding prediction cash to be allocated of sales counter can be used as the reality of cash time to be allocated by cash demand prediction model
Minimum cash needed for border.
Next propose the embodiment of the present application to the prebuild process of the cash demand prediction model in above-mentioned steps S102
It is illustrated, as shown in figure 3, being neural network training process schematic diagram provided by the embodiments of the present application, which mainly can be with
Include:
One, obtain multiple sample datas.
Wherein, each sample data may include: the sales counter geographic location of cash time to be allocated, cash to be allocated
Information, withdrawal affecting parameters, the sales counter history cash demand change information.
It is assumed that current time is on December 11st, 2018, the cash time to be allocated in each sample data be less than
Any time of current time.
The history cash demand change information of the sales counter of one sample data, the i.e. variation of cash demand in time domain scale
Parameter, optionally, history cash demand change information include: that the history in the previous preset time period of cash time to be allocated is existing
Subject to golden changes in demand information, select the peak of the actually required minimum cash of sales counter in the preset time period, minimum,
It is one or more in value, variance, root-mean-square value, flexure and kurtosis.For details, reference can be made to for testing data carry out description, this
In repeat no more.
The acquisition methods of sample data are illustrated below, the embodiment of the present application provides but is not limited to following methods:
Step 1 is based on the technologies such as Kafka, Flume, Sqoop and FTP, accesses in the corresponding data collection system of sales counter,
Acquiring in real time and the original cashes such as the cash transaction flowing water and cash transaction log that by way of batch capture, obtain the sales counter
Then transaction data (will be handed in collected original cash transaction data deposit HDFS original table because being stored with original cash
Easy data are so referred to as HDFS original table).
Step 2 carries out data cleansing and data mart modeling to the original cash transaction data of HDFS original table storage, obtains
Between cash transaction data.
Optionally, it can specifically include:
Remove noise data, extraneous data and repeated data etc..
Wherein, repeated data refers to the same transaction information of cash transaction flowing water and cash transaction log while record,
Or the repeated data of typing user withdrawal operation and generation is repeated as caused by mechanical disorder.
It is understood that there are users to reserve the case where wholesale is withdrawn the money, these situations are reserved in advance due to existing, so
It can prepare in advance, the data that these reservation wholesales are withdrawn the money will affect weight during trained neural network, optionally, unrelated number
According to referring to the wholesale withdrawal data reserved in advance.
It is understood that in technical dates, such as the original cash transaction data of electric business consumption day are more special, it is optional
, the original cash transaction data that electric business consumes day can be known as noise data.
Unstructured data and semi-structured data in intermediate cash transaction data are converted to structural data by step 3.
Wherein, structural data is that relevant database can be used come the number for the performance two dimensional form for indicating and storing
According to, can pass through intrinsic key assignments obtain corresponding information;Semi-structured data can be adjusted by flexible key assignments and obtain corresponding letter
Breath, but the format of data is not fixed;Unstructured data is the data of not fixed structure, the office text comprising full format
Shelves, text, picture, XML, HTML, all kinds of reports, image and audio/visual information etc..
Structural data is stored in HDFS increment list and full dose table by step 4, optionally, can be from the HDFS increment list
Sample data is obtained in full dose table.
Wherein, Kafka is the open source stream process platform developed by Apache Software Foundation, by Scala and Java
It writes.Kafka is that a kind of distributed post of high-throughput subscribes to message system.
Flume is the High Availabitity that Cloudera is provided, highly reliable, distributed massive logs acquisition, polymerization
With the system of transmission, Flume supports to customize Various types of data sender in log system, for collecting data;Meanwhile Flume
It provides and simple process is carried out to data, and write the ability of various data receivings (customizable).
Sqoop (pronunciation: skup) is the tool of a open source, is mainly used in Hadoop (Hive) and traditional database
Between (mysql, postgresql...) carry out data transmitting, can by a relevant database (such as: MySQL,
Oracle, Postgres etc.) in data lead in the HDFS for entering Hadoop, the data of HDFS can also be led and enter relationship type
In database.
FTP (File Transfer Protocol, File Transfer Protocol), pair of the control file on Internet
To transmission.Meanwhile it is also an application program (Application).
HDFS (Hadoop distributed file system) is designed to be suitble to operate in common hardware (commodity
Hardware the distributed file system on).
Two, using each sample data as the training input of neural network 31, training obtains the cash demand prediction
Model.
Optionally, great amount of samples data can be divided into several sample sets, each sample set may include multiple
Sample data.For example, 1000 sample datas are divided into 10 sample data sets, and to sample data sets label, each
Include 100 sample datas in set, neural network is updated respectively as unit of set.
Specifically, for neural network without manually given feature, it can learn from extensive sample data and extract spy
Sign, oneself is found to the better abstract expression method of sample data, it is hereby achieved that more better features.Neural network can
To obtain hundreds or even thousands kind feature, therefore, the accuracy that neural network handles sample data from a sample data
It is very high.
Neural network can use the right value update that back-propagation gradient descent algorithm carries out neural network, realize nerve net
The repetitive exercise of network and convergence.
It should be noted that neural network, which starts used weight, can be the random weight of initialization, neural network
First multiple sample datas are handled based on random weight, and are based on processing result, weight is updated;Then nerve net
Network is based on updated weight again and handles multiple sample datas;Weight is updated again based on processing result;Through
After crossing successive ignition, if the number of iterations is greater than preset times, or processing result meets termination condition deconditioning, obtains most
Whole cash demand prediction model.
Cash demand prediction model is to tend to the corresponding reality of sample data with the corresponding prediction cash to be allocated of sample data
Minimum cash needed for border is training objective, and training neural network obtains;So cash demand prediction model has the cabinet
The prediction cash to be allocated of platform tends to the ability of the actually required minimum cash of the sales counter, thus this method will predict it is to be allocated
Cash is as actually required minimum cash.
Neural network training process provided by the embodiments of the present application can specifically include:
Step 1: obtaining each sample data and respectively corresponding each sample data as the training input of neural network
Prediction cash to be allocated.
Optionally, when testing data is input to cash demand prediction model, advanced row data normalization is needed, then sample
Data are when being input to neural network, also advanced row data normalization;It should be noted that normalization is conducive to the fast of neural network
Speed convergence, it can accelerate weight stabilized speed;Optionally, cash demand prediction model output data is normalization data,
Need to carry out inverse normalization.
Optionally, neural network oneself itself has normalization and inverse normalized function, then sample data is inputting
It, can be without normalization, in the prediction cash to be allocated that exports neural network and actually required minimum existing when neural network
When gold is compared, without carrying out inverse normalization to prediction cash to be allocated.
Step 2: being directed to each sample data, the sample data corresponding prediction cash to be allocated and practical institute are obtained
Minimum cash comparison result is needed, to obtain the corresponding comparison result of each sample data.
Wherein, the actually required minimum cash of each sample data acceptance of the bid label characterizing part is minimum inventories cash, by step
Each sample data obtained in one is corresponding to predict that cash to be allocated is compared with the minimum inventories cash, obtains the sample set
The corresponding comparison result of each sample data for including in conjunction.
Step 3: being based on the corresponding comparison result of each sample data, the weight in the neural network is updated.
Optionally, as unit of sample data sets, based on each sample data for including in the sample data sets point
Not corresponding comparison result is updated the weighting parameter for including in each hidden layer of neural network.
Step 4: return step one, until the corresponding comparison result of each sample data meets termination condition;Obtain institute
State cash demand prediction model.
Optionally, return step one, using another sample set in sample data as the defeated of updated neural network
Enter, input each sample data in the set respectively, output obtain the corresponding prediction of all sample datas in the sample set to
Cash is distributed, step 2 and step 3 are repeated, obtains repeating updated neural network;Such as the sample data in examples detailed above
Including 10 sample data sets, and 10 sample data sets include: sample data sets 1, sample data sets 2, sample
Sample data sets 1 are input to cash demand prediction model, by step 3 by data acquisition system 3 ..., sample data sets 10
After obtaining updated neural network, sample data sets 2 are input to cash demand prediction model by return step one, are passed through
Step 3 obtains updated neural network again, and so on, until meeting termination condition.
Optionally, if the comparison result in step 4 meets termination condition (for example, the number of iterations is greater than or equal to default change
Generation number, or, the sum of corresponding comparison result of all sample datas is less than or equal to preset threshold), then neural network has been
It is trained to finish.
It should be noted that as time goes by, available new sample data can be based on new sample data again
It is secondary that cash demand prediction model is trained, to update the weight in cash demand prediction model.
Optionally, can be for the sample data in special day (such as electric business consumes day), individually training neural network obtains
To a cash demand prediction model, the sample data in special day includes: the sales counter of cash time to be allocated, cash to be allocated
Geographic location information, withdrawal affecting parameters, the sales counter history cash demand change information, sample data is retouched
It states, reference can be made to not repeated here for the explanation of sample data in step 1 in process shown in Fig. 3.
Based on any of the above-described embodiment, optionally, the embodiment of the present application also provides the following contents: at least based on described pre-
Cash to be allocated is surveyed, determines and adds the paper money frequency, and/or, add the paper money time.
If the cash demand prediction model referred in cash demand amount prediction technique shown in FIG. 1 is to include by a site
All sales counters corresponding sample data training neural network obtain;If sales counter refers to ATM machine, then, the ATM of a site
Machine predicts cash demand total amount=cash demand prediction model output prediction cash to be allocated;If sales counter refers to artificial sales counter,
So, artificial sales counter prediction cash demand total amount=cash demand prediction model output prediction of a site is to be allocated existing
Gold.
If the cash demand prediction model referred in cash demand amount prediction technique shown in FIG. 1 is corresponding by a sales counter
Sample data training neural network obtain;If sales counter refers to ATM machine, then, the ATM machine of a site predicts cash demand
Total amount=ATM1 prediction cash to be allocated+...+ATMr predicts cash to be allocated, and wherein r is the ATM machine sum that the site includes
Amount;If sales counter refers to artificial sales counter, then, artificial sales counter prediction cash demand total amount=artificial sales counter 1 of a site is predicted
Cash to be allocated+...+artificial sales counter m prediction cash to be allocated, wherein m is the artificial sales counter total quantity that the site includes.
One site prediction cash demand total amount=ATM machine prediction cash demand total amount+artificial sales counter predicts cash demand
Total amount.
Optionally, it if ATM machine prediction cash demand total amount is greater than 0, executes plus paper money, general execution when usual ATM clear paper money
Add paper money, in which:
ATM machine adds paper money number=ATM machine prediction cash demand total amount;
ATM machine adds the paper money frequency=round up (ATM adds paper money number/ATM paper money case quota)
Artificial sales counter adds paper money number=artificial sales counter prediction cash demand total amount;
Artificial sales counter adds the paper money frequency=round up (ATM adds paper money number/ATM paper money case quota).
Optionally, it is further contemplated that sales counter closing balance, to determine whether plus paper money.
If sales counter is ATM machine, by taking above-mentioned bank outlets include r platform ATM machine as an example, then, the overall balance of ATM machine is r platform ATM
The summation of machine remaining sum, it may be assumed that
ATM machine overall balance=ATM1 remaining sum+...+ATMr remaining sum.
If ATM machine predicts that cash demand total amount is more than or equal to ATM machine overall balance, execute plus paper money, one when the clear paper money of usual ATM
As execute plus paper money, in which:
ATM machine adds paper money number=ATM machine prediction cash demand total amount-ATM machine overall balance;
ATM machine adds the paper money frequency=round up (ATM adds paper money number/ATM paper money case quota).For example, ATM machine add paper money number=
ATM machine predicts cash demand total amount-ATM machine overall balance, r=3, and every ATM paper money case quota is 400,000,3 ATM machine overall balances
It is 800,000, ATM machine predicts that cash demand total amount is 1,500,000, since ATM machine plus paper money number are greater than 0, determine plus paper money, ATM machine add paper money
Number is 1,500,000-80 ten thousand=700,000, and ATM machine adds the paper money frequency to be (70/40)=2 time that round up.
Based on any of the above-described embodiment, optionally, the embodiment of the present application also provides the following contents: at least based on described pre-
Cash to be allocated is surveyed, determines the clear paper money frequency, and/or, the clear paper money time.
Optionally, determine that clear paper money time mode is as follows.
If ATM machine predicts that cash demand total amount is less than ATM machine overall balance, clear paper money may be needed.Cash is predicted in ATM machine
In the case that total demand is less than ATM machine overall balance, there are following several situations:
The first, ATM machine predicts that cash demand total amount is negative, illustrate that the credit of ATM machine is greater than withdrawal number,
And ATM paper money case quota is less than or equal to (ATM machine overall balance-ATM machine predicts cash demand total amount).For example, certain bank outlets wraps
Include three ATM machine, every ATM paper money case quota is 400,000, and ATM machine overall balance is 800,000, ATM machine predict cash demand total amount be-
500000, at this point, determining clear paper money.
Second, ATM machine predicts that cash demand total amount is less than ATM machine overall balance, and ATM paper money case quota is greater than (ATM machine is total
Remaining sum-ATM machine predicts cash demand total amount), at this point, optionally, it can also clear paper money.Optionally, in the latter case, although
ATM machine has " idle cash ", but in order not to waste clear paper money cost, maximum clear k days paper money period, such as 7 days (k of longest can be set
=7) necessary clear paper money, it is assumed that the ATM machine prediction cash demand total amount of prediction the 1st day is prediction cash demand amount 1 ..., prediction kth
It ATM machine prediction cash demand total amount is prediction cash demand amount k.At this point, ATM machine overall balance is greater than or equal to, (prediction is existing
Golden demand 1+...+ predicts cash demand amount t), wherein t=k+1, and:
The clear paper money frequency=be rounded ((current period last day ATM overall balance-first day next period prediction cash downwards
Demand)/ATM paper money case quota)
Such as certain bank outlets includes three ATM machine, it is 80 that every ATM paper money case quota, which is 400,000,3 ATM machine overall balances,
Ten thousand, if the maximum clear paper money period is 7 days, first day ATM machine prediction cash demand total amount is 600,000 in certain period, at this time without
Clear paper money, latter six days prediction cash demand total amounts are respectively -10 ten thousand, 200,000, -20 ten thousand, -30 ten thousand, 400,000,100,000, it is clear that herein
Daily ATM machine remaining sum is greater than second day prediction cash demand total amount in period, without clear paper money, if the 8th day ATM machine at this time
Predict that cash demand total amount is 50,000, then at this point, ATM machine prediction overall balance 800,000 is more than or equal to 750,000 (60-10 ten thousand+20 ten thousand -20
Ten thousand) ten thousand -30 ten thousand+40 ten thousand+10 ten thousand+5=75, due to reaching the maximum clear paper money period, then carry out clear paper money, in which:
Overall balance=100,000 current period last day ATM (800,000-60,+10 ten thousand-20 ten thousand+20 ten thousand+30 ten thousand-40 ten thousand-10 ten thousand
=10 ten thousand), and first day next period prediction cash demand amount is 50,000, then:
The clear paper money frequency=rounding ((100,000-5 ten thousand)/400,000) downwards=0 time, i.e., operate without clear paper money.
In another embodiment, above-mentioned 8th day ATM machine prediction cash demand total amount is -50 ten thousand, then:
The clear paper money frequency=rounding ((100,000+50 ten thousand)/400,000) downwards=1 time carries out primary clear paper money operation.
It is understood that be referred to above-mentioned ATM machine adds paper money frequency when cash sales counter to be allocated is people's work sales counter
Secondary, the clear paper money frequency plus paper money time and clear paper money time, the embodiment of the present application do not repeat them here.
Method is described in detail in aforementioned present invention disclosed embodiment, diversified forms can be used for method of the invention
Device realize that therefore the invention also discloses a kind of devices, and specific embodiment is given below and is described in detail.
As shown in figure 4, Fig. 4 is a kind of cash demand amount prediction meanss structural schematic diagram disclosed in the embodiment of the present application, it should
Device may include:
First obtains module, and for obtaining testing data, the testing data includes: cash time to be allocated, to be allocated
The sales counter geographic location information of cash, withdrawal affecting parameters, the sales counter history cash demand change information;
Determining module, for passing through the cash for the cash demand prediction model of testing data input prebuild
Demand Forecast Model determines the corresponding prediction cash to be allocated of the sales counter;
Wherein, the cash demand prediction model, which has, tends to the sales counter for the prediction cash to be allocated of the sales counter
The ability of actually required minimum cash.
Optionally, sales counter includes artificial sales counter or automatic sales counter, and cash demand amount prediction meanss can also include:
Second obtains module, for obtaining the corresponding multiple sample datas of ATM machine;
Third obtains module, and for the training input using each sample data as neural network, training obtains ATM machine
The corresponding cash demand prediction model.
Optionally, third acquisition module may include:
First acquisition unit obtains each sample number for the training input using each sample data as neural network
According to corresponding prediction cash to be allocated;
Second acquisition unit, for be directed to each sample data, obtain the sample data it is corresponding predict cash to be divided with
And actually required minimum cash comparison result, to obtain the corresponding comparison result of each sample data;
Updating unit updates the power in the neural network for being based on the corresponding comparison result of each sample data
Value parameter;
Trigger unit, for triggering first acquisition unit, until the corresponding comparison result of each sample data meets eventually
Only condition;Obtain the cash demand prediction model.
As shown in figure 5, being a kind of implementation structure chart of electronic equipment provided by the embodiments of the present application, electronic equipment packet
It includes:
Memory 501, for storing program;
Processor 502, for executing described program, described program is specifically used for:
Obtain testing data, the testing data include: cash time to be allocated, cash to be allocated sales counter where it is geographical
Location information, withdrawal affecting parameters, the sales counter history cash demand change information;
It is true by the cash demand prediction model by the cash demand prediction model of testing data input prebuild
The fixed sales counter is corresponding to predict cash to be allocated;
Wherein, the cash demand prediction model, which has, tends to the sales counter for the prediction cash to be allocated of the sales counter
The ability of actually required minimum cash.
Electronic equipment further include: bus, communication interface 503, input equipment 504 and output equipment 505.
Processor 502, memory 501, communication interface 503, input equipment 504 and output equipment 505 are mutual by bus
Connection.
The embodiment of the present application also provides a kind of readable storage medium storing program for executing, are stored thereon with computer program, described program tool
Body is used for:
Obtain testing data, the testing data include: cash time to be allocated, cash to be allocated sales counter where it is geographical
Location information, withdrawal affecting parameters, the sales counter history cash demand change information;
It is true by the cash demand prediction model by the cash demand prediction model of testing data input prebuild
The fixed sales counter is corresponding to predict cash to be allocated;
Wherein, the cash demand prediction model, which has, tends to the sales counter for the prediction cash to be allocated of the sales counter
The ability of actually required minimum cash.
It should be noted that all the embodiments in this specification are described in a progressive manner, each embodiment weight
Point explanation is the difference from other embodiments, and the same or similar parts between the embodiments can be referred to each other.
For device or system class embodiment, since it is basically similar to the method embodiment, so be described relatively simple, it is related
Place illustrates referring to the part of embodiment of the method.
It should also be noted that, herein, relational terms such as first and second and the like are used merely to one
Entity or operation are distinguished with another entity or operation, without necessarily requiring or implying between these entities or operation
There are any actual relationship or orders.Moreover, the terms "include", "comprise" or its any other variant are intended to contain
Lid non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (10)
1. a kind of cash demand amount prediction technique characterized by comprising
Testing data is obtained, the testing data includes: the sales counter geographic location of cash time to be allocated, cash to be allocated
Information, withdrawal affecting parameters, the sales counter history cash demand change information;
By the cash demand prediction model of testing data input prebuild, institute is determined by the cash demand prediction model
State the corresponding prediction cash to be allocated of sales counter;
Wherein, the cash demand prediction model has the reality that the prediction cash to be allocated of the sales counter is tended to the sales counter
The ability of required minimum cash.
2. cash demand amount prediction technique according to claim 1, which is characterized in that the sales counter is for artificial sales counter or automatically
Sales counter, further includes:
Obtain multiple sample datas;
Using each sample data as the training input of neural network, training obtains the cash demand prediction model.
3. cash demand amount prediction technique according to claim 1 or claim 2, which is characterized in that described to obtain each sample data work
For the training input of neural network, training obtains the cash demand prediction model and includes:
Using each sample data as the training input of neural network, it is to be allocated to obtain the corresponding prediction of each sample data
Cash;
For each sample data, obtain that the sample data is corresponding to predict that cash to be divided and actually required minimum cash compare
As a result, to obtain the corresponding comparison result of each sample data;
Based on the corresponding comparison result of each sample data, the weighting parameter in the neural network is updated;
Training input of the return step using each sample data as neural network, it is corresponding pre- to obtain each sample data
Survey cash to be allocated;Until the corresponding comparison result of each sample data meets termination condition;It is pre- to obtain the cash demand
Survey model.
4. cash demand amount prediction technique according to claim 3, which is characterized in that
The corresponding actually required minimum cash of each sample data is minimum inventories cash.
5. cash demand amount prediction technique according to claim 1, which is characterized in that further include:
At least based on prediction cash to be allocated, determine plus the paper money frequency, and/or, the clear paper money frequency, and/or, add the paper money time, and/
Or, the clear paper money time.
6. a kind of cash demand amount prediction meanss characterized by comprising
First obtains module, and for obtaining testing data, the testing data includes: cash time to be allocated, cash to be allocated
Sales counter geographic location information, withdrawal affecting parameters, the sales counter history cash demand change information;
Determining module, for passing through the cash demand for the cash demand prediction model of testing data input prebuild
Prediction model determines the corresponding prediction cash to be allocated of the sales counter;
Wherein, the cash demand prediction model has the reality that the prediction cash to be allocated of the sales counter is tended to the sales counter
The ability of required minimum cash.
7. cash demand amount prediction meanss according to claim 6, which is characterized in that the sales counter include artificial sales counter or from
Dynamic sales counter, further includes:
Second obtains module, for obtaining the corresponding multiple sample datas of ATM machine;
Third obtains module, and for the training input using each sample data as neural network, it is corresponding that training obtains ATM machine
The cash demand prediction model.
8. cash demand amount prediction meanss according to claim 6, which is characterized in that the third obtains module and includes:
First acquisition unit obtains each sample data point for the training input using each sample data as neural network
Not corresponding prediction cash to be allocated;
Second acquisition unit, for obtaining for each sample data, the sample data is corresponding to predict cash to be divided and reality
Minimum cash comparison result needed for border, to obtain the corresponding comparison result of each sample data;
Updating unit updates the weight ginseng in the neural network for being based on the corresponding comparison result of each sample data
Number;
Trigger unit, for triggering first acquisition unit, until the corresponding comparison result of each sample data, which meets, terminates item
Part;Obtain the cash demand prediction model.
9. a kind of electronic equipment characterized by comprising
Memory, for storing program;
Processor, for executing described program, described program is specifically used for:
Testing data is obtained, the testing data includes: the sales counter geographic location of cash time to be allocated, cash to be allocated
Information, withdrawal affecting parameters, the sales counter history cash demand change information;
By the cash demand prediction model of testing data input prebuild, institute is determined by the cash demand prediction model
State the corresponding prediction cash to be allocated of sales counter;
Wherein, the cash demand prediction model has the reality that the prediction cash to be allocated of the sales counter is tended to the sales counter
The ability of required minimum cash.
10. a kind of readable storage medium storing program for executing, is stored thereon with computer program, which is characterized in that the computer program is processed
When device executes, each step that cash demand amount prediction technique as claimed in claim 1 to 5 includes is realized.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910108630.3A CN109829818A (en) | 2019-02-03 | 2019-02-03 | Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910108630.3A CN109829818A (en) | 2019-02-03 | 2019-02-03 | Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109829818A true CN109829818A (en) | 2019-05-31 |
Family
ID=66863463
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910108630.3A Pending CN109829818A (en) | 2019-02-03 | 2019-02-03 | Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109829818A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110400022A (en) * | 2019-07-31 | 2019-11-01 | 中国工商银行股份有限公司 | Self-help teller machine cash dosage prediction technique and device |
CN110400021A (en) * | 2019-07-31 | 2019-11-01 | 中国工商银行股份有限公司 | Bank outlets' cash dosage prediction technique and device |
CN110415462A (en) * | 2019-07-31 | 2019-11-05 | 中国工商银行股份有限公司 | Atm device adds paper money optimization method and device |
CN110659825A (en) * | 2019-09-23 | 2020-01-07 | 中国银行股份有限公司 | Cash demand prediction method and device for multiple learners of bank outlets |
CN111145015A (en) * | 2019-12-31 | 2020-05-12 | 中国银行股份有限公司 | Reservation-free withdrawal amount determining method and device |
CN111179065A (en) * | 2019-12-31 | 2020-05-19 | 中国银行股份有限公司 | Reservation-free withdrawal amount determining method and device |
CN111738504A (en) * | 2020-06-19 | 2020-10-02 | 中国工商银行股份有限公司 | Enterprise financial index fund amount prediction method and device, equipment and storage medium |
CN111754325A (en) * | 2020-06-24 | 2020-10-09 | 中国银行股份有限公司 | Service data processing method and system |
CN111986406A (en) * | 2020-09-03 | 2020-11-24 | 中国银行股份有限公司 | Data processing method and device based on withdrawal transaction |
CN113222512A (en) * | 2021-05-24 | 2021-08-06 | 中国农业银行股份有限公司 | Data processing method and device of self-service equipment and equipment |
CN113449103A (en) * | 2021-01-28 | 2021-09-28 | 民生科技有限责任公司 | Bank transaction flow classification method and system integrating label and text interaction mechanism |
CN113988461A (en) * | 2021-11-11 | 2022-01-28 | 中国工商银行股份有限公司 | Position prediction method, position prediction device, storage medium and electronic equipment |
CN114495378A (en) * | 2022-01-21 | 2022-05-13 | 浪潮卓数大数据产业发展有限公司 | Cash withdrawal information acquisition and processing method and system based on ATM |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104424598A (en) * | 2013-09-06 | 2015-03-18 | 株式会社日立制作所 | Cash demand quantity predicating device and method |
CN106886846A (en) * | 2017-04-26 | 2017-06-23 | 中南大学 | A kind of bank outlets' excess reserve Forecasting Methodology that Recognition with Recurrent Neural Network is remembered based on shot and long term |
CN106952420A (en) * | 2016-01-07 | 2017-07-14 | 株式会社日立制作所 | ATM cash management device, system and method |
CN107316111A (en) * | 2017-07-04 | 2017-11-03 | 深圳信用宝金融服务有限公司 | A kind of withdraw deposit Forecasting Methodology and device based on internet |
CN107909463A (en) * | 2017-10-26 | 2018-04-13 | 交通银行股份有限公司 | Bank outlets' cash demand Forecasting Methodology and device, prediction allot system |
-
2019
- 2019-02-03 CN CN201910108630.3A patent/CN109829818A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104424598A (en) * | 2013-09-06 | 2015-03-18 | 株式会社日立制作所 | Cash demand quantity predicating device and method |
CN106952420A (en) * | 2016-01-07 | 2017-07-14 | 株式会社日立制作所 | ATM cash management device, system and method |
CN106886846A (en) * | 2017-04-26 | 2017-06-23 | 中南大学 | A kind of bank outlets' excess reserve Forecasting Methodology that Recognition with Recurrent Neural Network is remembered based on shot and long term |
CN107316111A (en) * | 2017-07-04 | 2017-11-03 | 深圳信用宝金融服务有限公司 | A kind of withdraw deposit Forecasting Methodology and device based on internet |
CN107909463A (en) * | 2017-10-26 | 2018-04-13 | 交通银行股份有限公司 | Bank outlets' cash demand Forecasting Methodology and device, prediction allot system |
Non-Patent Citations (2)
Title |
---|
张晓明: "《商业银行经营管理》", 31 March 2012, 清华大学出版社 北京交通大学出版社 * |
蒋理,马超群: "《中国制造2025智能制造企业信息系统》", 31 May 2018, 湖南大学出版社 * |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110400022A (en) * | 2019-07-31 | 2019-11-01 | 中国工商银行股份有限公司 | Self-help teller machine cash dosage prediction technique and device |
CN110400021A (en) * | 2019-07-31 | 2019-11-01 | 中国工商银行股份有限公司 | Bank outlets' cash dosage prediction technique and device |
CN110415462A (en) * | 2019-07-31 | 2019-11-05 | 中国工商银行股份有限公司 | Atm device adds paper money optimization method and device |
CN110400021B (en) * | 2019-07-31 | 2022-03-25 | 中国工商银行股份有限公司 | Bank branch cash usage prediction method and device |
CN110400022B (en) * | 2019-07-31 | 2022-03-04 | 中国工商银行股份有限公司 | Cash consumption prediction method and device for self-service teller machine |
CN110659825A (en) * | 2019-09-23 | 2020-01-07 | 中国银行股份有限公司 | Cash demand prediction method and device for multiple learners of bank outlets |
CN111145015A (en) * | 2019-12-31 | 2020-05-12 | 中国银行股份有限公司 | Reservation-free withdrawal amount determining method and device |
CN111179065A (en) * | 2019-12-31 | 2020-05-19 | 中国银行股份有限公司 | Reservation-free withdrawal amount determining method and device |
CN111179065B (en) * | 2019-12-31 | 2023-04-18 | 中国银行股份有限公司 | Reservation-free withdrawal amount determining method and device |
CN111145015B (en) * | 2019-12-31 | 2023-06-23 | 中国银行股份有限公司 | Reservation-free withdrawal amount determining method and device |
CN111738504A (en) * | 2020-06-19 | 2020-10-02 | 中国工商银行股份有限公司 | Enterprise financial index fund amount prediction method and device, equipment and storage medium |
CN111754325B (en) * | 2020-06-24 | 2023-09-12 | 中国银行股份有限公司 | Service data processing method and system |
CN111754325A (en) * | 2020-06-24 | 2020-10-09 | 中国银行股份有限公司 | Service data processing method and system |
CN111986406B (en) * | 2020-09-03 | 2022-02-15 | 中国银行股份有限公司 | Data processing method and device based on withdrawal transaction |
CN111986406A (en) * | 2020-09-03 | 2020-11-24 | 中国银行股份有限公司 | Data processing method and device based on withdrawal transaction |
CN113449103A (en) * | 2021-01-28 | 2021-09-28 | 民生科技有限责任公司 | Bank transaction flow classification method and system integrating label and text interaction mechanism |
CN113449103B (en) * | 2021-01-28 | 2024-05-10 | 民生科技有限责任公司 | Bank transaction running water classification method and system integrating label and text interaction mechanism |
CN113222512A (en) * | 2021-05-24 | 2021-08-06 | 中国农业银行股份有限公司 | Data processing method and device of self-service equipment and equipment |
CN113988461A (en) * | 2021-11-11 | 2022-01-28 | 中国工商银行股份有限公司 | Position prediction method, position prediction device, storage medium and electronic equipment |
CN114495378A (en) * | 2022-01-21 | 2022-05-13 | 浪潮卓数大数据产业发展有限公司 | Cash withdrawal information acquisition and processing method and system based on ATM |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109829818A (en) | Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing | |
CN106549772B (en) | Resource prediction method, system and capacity management device | |
Law et al. | A neural network model to forecast Japanese demand for travel to Hong Kong | |
CN110400022A (en) | Self-help teller machine cash dosage prediction technique and device | |
Sing et al. | Dynamic modeling of workforce planning for infrastructure projects | |
CN104424598A (en) | Cash demand quantity predicating device and method | |
CN108388974A (en) | Top-tier customer Optimum Identification Method and device based on random forest and decision tree | |
CN107506868A (en) | A kind of method and device of temporary electricity load prediction | |
CN109784959A (en) | A kind of target user's prediction technique, device, background server and storage medium | |
Feng et al. | [Retracted] Design and Simulation of Human Resource Allocation Model Based on Double‐Cycle Neural Network | |
CN110059052A (en) | A kind of refinery scheduling case management method and computer readable storage medium | |
CN108154311A (en) | Top-tier customer recognition methods and device based on random forest and decision tree | |
CN109685643A (en) | Loan audit risk grade determines method, apparatus, equipment and storage medium | |
CN106407305A (en) | Data mining system and method | |
Roach | Reconciled boosted models for GEFCom2017 hierarchical probabilistic load forecasting | |
CN108921425A (en) | A kind of method, system and the server of asset item classifcation of investment | |
Foong et al. | Power plant maintenance scheduling using ant colony optimization: an improved formulation | |
EP1212716A1 (en) | Interaction prediction system and method | |
CN106407316A (en) | Topic model-based software question and answer recommendation method and device | |
CN107977855A (en) | A kind of method and device of managing user information | |
Cai | Spatial differentiation of intangible cultural heritage in South China and its influencing factors | |
CN108629625A (en) | A kind of monthly electricity sales amount prediction technique, device and server | |
CN111523570A (en) | Smart city system based on community post house and control method thereof | |
CN109741172A (en) | Credit method for early warning, device, system and storage medium | |
CN110197316A (en) | Processing method, device, computer-readable medium and the electronic equipment of operation data |
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: 20190531 |
|
RJ01 | Rejection of invention patent application after publication |