CN113487110A - Spare payment management method and device - Google Patents
Spare payment management method and device Download PDFInfo
- Publication number
- CN113487110A CN113487110A CN202110856998.5A CN202110856998A CN113487110A CN 113487110 A CN113487110 A CN 113487110A CN 202110856998 A CN202110856998 A CN 202110856998A CN 113487110 A CN113487110 A CN 113487110A
- Authority
- CN
- China
- Prior art keywords
- index value
- model
- target
- autoregressive model
- evaluation index
- 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
Images
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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
Abstract
The invention discloses a spare payment management method and a spare payment management device, which relate to the technical field of finance, wherein the method comprises the following steps: acquiring daily throughput data of an intelligent counter of a bank outlet in a preset time period; constructing an autoregressive model by using preset parameters, training the autoregressive model by using the throughput data, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model; predicting the end balance of a specified date in the handling data by using an intermediate model to obtain a predicted end balance; calculating an evaluation index value of the intermediate model according to the throughput data and the predicted end balance; comparing the calculated evaluation index value with a preset standard index value range; and if the evaluation index value is in the standard index value range, determining the intermediate model as a target autoregressive model, and predicting the reserve payment single-day deposit estimate of the future target date by using the target autoregressive model. The invention can reduce the error probability caused by subjective prediction.
Description
Technical Field
The invention relates to the technical field of finance, in particular to a spare payment management method and device.
Background
This section is intended to provide a background or context to the embodiments of the invention that are recited in the claims. The description herein is not admitted to be prior art by inclusion in this section.
In the daily maintenance process, a manager needs to add cash to an intelligent counter in a website at a certain cash adding frequency and cash adding amount so as to ensure that sufficient cash is available in the intelligent counter for a cardholder to take.
The traditional spare payment management method is to roughly estimate the money adding amount according to the report sent on the intelligent counter of each network point and the management experience of cash managers, and has strong subjectivity and higher error probability.
Disclosure of Invention
The embodiment of the invention provides a method for managing reserve payment, which is used for intelligently predicting the single-day deposit estimated amount of the reserve payment through historical data, reducing the error probability caused by subjectivity, reducing the bank operation cost and improving the intelligent counter fund utilization efficiency, and comprises the following steps:
acquiring daily handling data of an intelligent counter of a bank outlet in a preset time period, wherein the handling data comprises deposit amount of each time of a deposit fund, withdrawal amount of each time, single-day end balance and corresponding recording time;
constructing an autoregressive model by using preset parameters, training the autoregressive model by using the throughput data, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model;
predicting the end balance of a specified date in the handling data by using an intermediate model to obtain a predicted end balance, wherein the specified date is one day of the recording time in the handling data;
calculating an evaluation index value of the intermediate model according to the throughput data and the predicted end balance;
comparing the calculated evaluation index value with a preset standard index value range;
and if the evaluation index value is in the standard index value range, determining the intermediate model as a target autoregressive model, and predicting the reserve payment single-day deposit estimate of the future target date by using the target autoregressive model.
The embodiment of the invention also provides a spare payment management device, which is used for intelligently predicting the single-day deposit estimated amount of the spare payment through historical data, reducing the error probability caused by subjectivity, reducing the bank operation cost and improving the intelligent counter fund utilization efficiency, and comprises the following steps:
the system comprises an acquisition module, a storage module and a management module, wherein the acquisition module is used for acquiring the daily handling data of an intelligent counter of a bank outlet in a preset time period, and the handling data comprises the deposit amount of each time of a deposit fund, the withdrawal amount of each time, the single-day end balance and the corresponding recording time;
the model training module is used for constructing an autoregressive model by using preset parameters, training the autoregressive model by using the throughput data, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model;
the forecasting module is used for forecasting the end balance of the appointed date in the handling data by utilizing the intermediate model to obtain the forecast end balance, wherein the appointed date is one day in the recording time in the handling data;
the calculation module is used for calculating an evaluation index value of the intermediate model according to the throughput data and the predicted end balance;
the comparison module is used for comparing the calculated evaluation index value with a preset standard index value range;
and the prediction module is used for determining the intermediate model as a target autoregressive model when the evaluation index value is within the standard index value range, and predicting the reserve fund single-day deposit estimate of the future target date by using the target autoregressive model.
The embodiment of the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the reserve payment management method when executing the computer program.
An embodiment of the present invention further provides a computer-readable storage medium, where a computer program for executing the method for managing a reserve payment is stored in the computer-readable storage medium.
In the embodiment of the invention, the throughput data of the intelligent counter in the preset time period is obtained, the throughput data and the preset parameters are utilized to establish the autoregressive model, and an evaluation index value is calculated through the prediction result and the throughput data of the autoregressive model to evaluate the accuracy of the established autoregressive model, when the evaluation index value is determined to be in the standard index value range, the accuracy of the autoregressive model is determined to reach the standard, the method can predict the future single-day deposit forecast of the future fund of a certain target date, compared with the prediction by the experience of the staff in the prior art, the method can mine the periodic rule of the historical throughout data, therefore, the suitable spare payment money adding amount is found, the bank operation cost is reduced, the cash utilization efficiency of the intelligent counter is improved, the operation cost of the cash carrier is reduced, and the problem of low customer experience caused by the failure of money taking is avoided.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts. In the drawings:
FIG. 1 is a flowchart of a method for managing funds on deposit according to an embodiment of the present invention;
FIG. 2 is a flowchart of an exemplary method of implementing step 102 in accordance with an embodiment of the present invention;
FIG. 3 is a general flowchart of a method for managing funds prepared in accordance with an embodiment of the present invention;
FIG. 4 is another flowchart of a method for managing funds in preparation for payment according to an embodiment of the present invention;
FIG. 5 is a flow chart of a process for performing a training set and test set centralized throughput data update in an embodiment of the present invention;
FIG. 6 is another flowchart of a method for managing funds in preparation for payment according to an embodiment of the present invention;
FIG. 7 is another flowchart of a method for managing funds in preparation for payment according to an embodiment of the present invention;
FIG. 8 is a graph of end balance versus time in an embodiment of the present invention;
FIG. 9 is a diagram illustrating a relationship between a late balance and time in a holiday elimination period according to an embodiment of the present invention;
FIG. 10 is a schematic structural diagram of a spare payment management apparatus according to an embodiment of the present invention;
fig. 11 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention, but not to limit the present invention.
The time series is a statistical technology for predicting the development trend of future data according to a data generation cycle based on historical data, and in the process of managing the reserve payment, the management requirements of aspects such as reducing the cost of machines and tools deposited by the reserve payment on an intelligent counter, reducing the transportation cost of a cash carrier, meeting the requirements of customers to the maximum extent and the like are originally reduced, and the embodiment of the invention provides a reserve payment management method, as shown in fig. 1, the method comprises the following steps 101 to 106:
The handling data comprises deposit amount of each time of the deposit fund, withdrawal amount of each time, single-day end balance and corresponding recording time.
Considering that the frequency and the amount of cash deposit and withdrawal in the intelligent counters of different bank outlets are greatly different due to geographical positions and the like, even the amount of cash deposit and withdrawal in different intelligent counters of the same bank outlet are sometimes different to a certain extent, in order to ensure the accuracy of prediction, the daily throughput data of each intelligent counter in a preset time period can be respectively obtained, and the predicted amount of deposit and withdrawal in each intelligent counter on a single day is respectively predicted.
The preset time period may be set manually, such as set to 200 days, 300 days, etc. Because the current single-day end balance of the intelligent counter needs to be read from the cash box by a network worker, the bank also has a holiday, and the single-day end balance of the intelligent counter on the holiday cannot be obtained, the handling data of the bank worker in a preset time period and each day of the single-day end balance can be read in consideration of actual conditions. In addition, because the invention focuses on how to obtain a high-accuracy prediction result by using a reasonable data volume, the existing method requires a large amount of data, and the larger the amount of data, the more the models can reflect different data characteristics, and the proper preset time period which can be slightly shorter is selected in the embodiment of the invention.
In addition, considering that there are many clients who may deposit and withdraw money on holidays compared with weekdays, in one implementation, throughput data recorded on holidays can be labeled to determine whether to train an autoregressive model together with throughput data on holidays and weekdays according to actual conditions of model construction.
The intelligent counter has a large number of card holders for depositing and withdrawing every day, and in the embodiment of the invention, the deposit and withdrawal amount of each time operated on the intelligent counter by the card holders is obtained. Meanwhile, in order to distinguish between deposit and withdrawal, the deposit amount may be recorded as positive and the withdrawal amount may be recorded as negative.
Besides data such as single-day end balance and the like, the handling data can also comprise data such as deposit of each time, the occurrence time of withdrawal, machine numbers of intelligent counters and the like.
102, constructing an autoregressive model by using preset parameters, training the autoregressive model by using throughput data, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model.
The inventor finds through experimental study that a periodic relationship exists between the change amount of the bill amount per day and the single-day end balance, for example, taking 60 days as a period as an example, the single-day end balance of a day after 60 days can be predicted through the change amount of the bill amount of the day before 60 days, so that a period with a large influence on the end balance is selected as a preset parameter to establish a proper autoregressive model with high accuracy.
The autoregressive model is a statistical one that is commonly used in processing time series. In the process of determining the utilization of the autoregressive model, the inventor researches and researches a technology of spare payment management at home and abroad, considers the spare payment prediction problem as a curve trend prediction problem, further widens the problem research surface, innovatively proposes to research the spare payment management problem by utilizing the autoregressive model, and corrects the traditional autoregressive model according to the spare payment research characteristics and the small-scale test results of several rounds of models. The specific process of the model correction experiment is detailed in the subsequent content of the specification.
Specifically, as shown in fig. 2, step 102 may be executed as following steps 1021 to 1023:
And 1022, dividing the throughput data into a training set and a test set according to a preset proportion according to the sequence of the recording time.
The recording time of the throughput data in the training set is prior to the recording time of the throughput data in the testing set, and the preset proportion is used for indicating the proportion of the data volume in the testing set and the data volume in the training set to the total quantity of the throughput data respectively.
And 1023, training the autoregressive model with the preset parameters by using the training set, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model.
The preset proportion can be set manually, and the preset proportion is set to be the proportion which can achieve better model training effect according to experience of a training model, for example, the preset proportion can be set to be 4: 1. For example, if 200 days of throughput data are acquired, and the preset ratio is 4:1, the throughput data recorded in the first 160 days is divided into training sets, and the throughput data recorded in the last 40 days is divided into test sets.
The training set is used for training the autoregressive model, the testing set is used for testing the trained autoregressive model, the training and the testing are combined, the accuracy of the autoregressive model can be improved, and the test set verifies that the autoregressive model has high feasibility in application.
And 103, predicting the end balance of the appointed date in the throughput data by using the intermediate model to obtain the predicted end balance.
Wherein the specified date is one of the recording times in the throughput data. For example, if the throughput data recording time is from 6 months 1 to 12 months 31, one day may be arbitrarily selected as the specified date from 6 months 1 to 12 months 31, such as 6 months 15 or 12 months 6.
And 104, calculating an evaluation index value of the intermediate model according to the throughput data and the predicted end-of-term balance.
And 105, comparing the calculated evaluation index value with a preset standard index value range.
In order to avoid overfitting of the autoregressive model and consider the persuasion of the model, evaluation indexes are set to evaluate the intermediate model in the embodiment of the invention. The evaluation indexes are specifically coefficient Multiple R of linear regression, fitting coefficient R Square, Adjusted fitting coefficient Adjusted R Square, standard error and observed value. Since the evaluation index is an evaluation index commonly used in the prior art, a specific calculation method thereof can be referred to in the prior art, and details are not repeated in the embodiment of the present invention.
The standard index range is set in advance with respect to the accuracy of the established autoregressive model by human reference.
And step 106, if the evaluation index value is in the standard index value range, determining the intermediate model as a target autoregressive model, and predicting the reserve payment single-day deposit estimate of the future target date by using the target autoregressive model.
The target date single day end balance, deposit amount and withdrawal amount can be predicted according to the autoregressive model, and then the single day deposit estimated amount of the deposit fund is determined according to the single day end balance, the deposit amount and the withdrawal amount.
In the embodiment of the invention, if the evaluation index value is not in the standard index value range, the preset parameters are adjusted, the autoregressive model is reconstructed, the reconstructed autoregressive model is trained by using throughput data until the evaluation index value determined according to the reconstructed autoregressive model is in the standard index value range.
In another implementation mode, if the preset parameters are adjusted for multiple times, the reconstructed autoregressive model still cannot meet the condition that the evaluation index value is within the standard index value range, the information is fed back to research and development and maintenance personnel, so that the relevant personnel can reasonably solve the abnormal event.
Due to the fact that the use condition of the intelligent counter of the bank outlet changes with time, the autoregressive model built according to historical throughput data is not applicable after being used for a period of time, and therefore the model needs to be maintained.
Specifically, after the future fund single-day deposit estimate of the target date is predicted by using the target autoregressive model, the handling data of the target date can be acquired when the target date is finished; adding the handling data of the target date into the test set, selecting the handling data with the same volume as the handling data of the target date from the test set according to the sequence of the recording time from first to last, removing the test set and moving the test set into the training set, selecting the handling data with the same volume as the handling data of the target date from the training set and moving the training set out to obtain an updated training set and test set.
Referring to fig. 6, after acquiring the throughput data of the target date and updating the training set and the test set each time, calculating an evaluation index value of the currently used target regression model by using the updated training set, the updated test set, and the predicted end-of-term balance of the target date; and if the evaluation index value of the currently used target regression model is within the standard index value range, continuing to use the target autoregressive model, updating the training set and the test set every day until the time length for using the target autoregressive model reaches the maintenance date, adjusting the preset parameters, and re-determining the target autoregressive model by using the updated training set and the test set.
Typically, the target date is the day after the current time, for example, 7 months and 5 days today, and the autoregressive model is used today to predict the single-day placement estimate for the future payment of 7 months and 6 days. After 7/month and 6/day, actual throughput data for 7/month and 6/day can be collected. The 7-month 6-day throughput data has been used as historical data to update the training set and test set. Specifically, if the recording time of the throughput data in the currently used training set is from 2 months 5 days to 6 months 4 days, and the recording time of the throughput data in the used testing set is from 6 months 5 days to 7 months 4 days, the throughput data of 7 months and 5 days is added into the testing set, the throughput data of 6 months and 5 days is removed from the testing set, the throughput data of 6 months and 5 days is added into the training set, and the throughput data of 2 months and 5 days in the training set is removed at the same time, so that the data in the training set and the testing set are ensured to follow a preset proportion while the throughput data in the training set and the testing set are updated. The process of performing the throughput data update in the training set and the test set can be seen in fig. 5.
In the embodiment of the invention, a maintenance period is set, for example, 30 days, if the evaluation index values of the autoregressive model are within the standard index value range in the maintenance period, the autoregressive model is not changed and is still used continuously until the maintenance period is reached, and the autoregressive model is retrained by using the updated training set and the test set; if the evaluation index value of the autoregressive model exceeds the standard index value range within 30 days of the maintenance cycle, the preset parameters are adjusted, or the data volume of given throughput data is adjusted, such as changing the data of 200 days into the data of 300 days, and the autoregressive model is retrained to ensure the usability of the autoregressive model.
After calculating the evaluation index value of the currently used target regression model by using the updated training set, the updated test set and the predicted end-of-term balance of the target date, if the evaluation index value of the currently used target regression model is not within the standard index value range, as shown in fig. 7, adjusting the preset parameters, and re-determining the target autoregressive model by using the updated training set and the updated test set.
For easy understanding, the embodiment of the present invention further provides a general flowchart of a method for managing a fund, as shown in fig. 3.
In another implementation, referring to fig. 4, after the auto-regression model with preset parameters is trained by using a training set, and other parameters except the preset parameters in all undetermined parameters of the auto-regression model are determined to obtain an intermediate model, the end balance of the first specified date in the training set can be predicted by using the intermediate model to obtain a first predicted end balance; calculating a first evaluation index value of the intermediate model according to the training set and the first prediction end balance; comparing the first evaluation index value obtained by calculation with a preset standard index value range; if the first evaluation index value is within the standard index value range, predicting the end balance of a second specified date in the test set by using an intermediate model to obtain a second predicted end balance; calculating a second evaluation index value of the intermediate model according to the test set and the second prediction end balance; comparing the calculated second evaluation index value with a preset standard index value range; and if the second evaluation index value is within the standard index value range, determining the intermediate model as a target autoregressive model, and predicting the deposit amount of the reserve payment on the target date by using the target autoregressive model.
In the embodiment of the invention, the throughput data of the intelligent counter in the preset time period is obtained, the throughput data and the preset parameters are utilized to establish the autoregressive model, and an evaluation index value is calculated through the prediction result and the throughput data of the autoregressive model to evaluate the accuracy of the established autoregressive model, when the evaluation index value is determined to be in the standard index value range, the accuracy of the autoregressive model is determined to reach the standard, the method can predict the future single-day deposit forecast of the future fund of a certain target date, compared with the prediction by the experience of the staff in the prior art, the method can mine the periodic rule of the historical throughout data, therefore, the suitable spare payment money adding amount is found, the bank operation cost is reduced, the cash utilization efficiency of the intelligent counter is improved, the operation cost of the cash carrier is reduced, and the problem of low customer experience caused by the failure of money taking is avoided.
To demonstrate the feasibility of the autoregressive model, the experimental procedure for performing feasibility experiments on the basis of the autoregressive model is briefly described below.
One, experiment one
And (4) screening out the banknote quantity change data with the mechanism number of 51618 for later use through an Excel screening function. And selecting the end balance, namely the end balance of the money box as a standby variable. And (3) combining the characteristics of the autoregressive model, taking the last day balance as an independent variable and taking the last day balance as a dependent variable, and performing unary autoregressive analysis to obtain the analysis results of the following table I and table II:
watch 1
Watch two
Unary autoregressive non-intercept
From the analysis results, whether the model containing the intercept term (corresponding to table one) or the model containing no intercept term (corresponding to table two), the variable can be interpreted by the model in a very small part, so the traditional autoregressive model is not suitable for the experiment.
Second, experiment two
And correcting the traditional autoregressive model, and realizing significance of the returned quantity of each cash box in the deep-digging and paying operation. By consulting the data and interviewing with the cash deep manager, the fact that the change amount of the money amount and the balance at the end of the money amount each day may have a correlation is found, so that the change amount of the money amount on the previous day is selected as an independent variable, and the balance at the end of the money amount on the day is selected as a dependent variable to perform regression analysis, and the following three conclusions are obtained:
watch III
Unary increment non-intercept
In the research process, the default intercept item of the regression model is found to be an inherent banknote quantity of the banknote box corresponding to the problem of the reserve payment, and in the actual banknote box management, the banknote box does not have an inherent banknote quantity, so that in the process of researching the problem of the reserve payment, the default intercept item is the non-intercept model.
As can be seen from the regression analysis results, the explanation degree of the model is greatly improved, and the research direction of the model can be ensured to be correct.
Third, experiment three
Based on the research experience of conventional time series analysis, we can confirm a period that has a significant effect on the model results. In the second experiment, the banknote quantity change quantity of the previous day is selected as the independent variable according to the inherent mode of the model design, and in fact, a period can be found completely, and the day or days with the most obvious influence of the banknote quantity change quantity on the balance at the end of the current day are selected as the model independent variable according to the period. During the set-up process, a graph of the end of the monetary amount balance versus time is made, as shown in fig. 8 and 9. Fig. 8 is a diagram showing a relationship between an end balance of a certain period and time, and fig. 9 is a diagram showing a relationship between an end balance of a date after holiday elimination of the period and time.
From the relationship between the end balance and time, it can be easily seen that the trend of the end balance curve changes periodically. The periodic relationship between the change amount of the bill amount and the balance at the end of the period can be obtained by applying spectrum analysis. And respectively taking out the two-day money amount change and the end balance which have the most obvious influence, and carrying out regression analysis on the three-day change and the end balance to obtain the analysis results of the following table I and table II, wherein the table IV is the analysis results of periods of 71 days and 74 days, and the table V is the analysis results of periods of 71 days, 74 days and 60 days.
Watch four
Watch five
It can be seen from the above analysis results that the selection of the cycle having the most influence on the end-of-term balance is of great help to improve the accuracy of the prediction regression data, and the more data periods the model introduces, the higher the interpretation degree of the data can be by the model.
In the above experiments, the model is linked with model test through model establishment, and finally, a model suitable for the subject is found through small-batch data test. In the process of establishing the previous model, a model frame is often set firstly, only debugging parameters are needed, and in the test process of the model, the independent variable dependent variable relation of the model is modified through communication with each party, so that the model with higher prediction accuracy is obtained. Therefore, the method has the characteristics that the method dares to be new in the process of debugging the model, and makes correction for improving the accuracy of model prediction based on the basic model.
The embodiment of the invention also provides a spare payment management device, which is described in the following embodiment. Because the principle of the device for solving the problems is similar to the method for managing the reserve payment, the implementation of the device can refer to the implementation of the method for managing the reserve payment, and repeated parts are not described again.
As shown in fig. 10, the apparatus 1000 includes an obtaining module 1001, a model training module 1002, a predicting module 1003, a calculating module 1004, and an aligning module 1005.
The acquiring module 1001 is used for acquiring daily handling data of an intelligent counter of a bank outlet in a preset time period, wherein the handling data comprises deposit amount of each deposit of a deposit fund, withdrawal amount of each withdrawal, single-day end balance and corresponding recording time;
the model training module 1002 is configured to construct an autoregressive model by using preset parameters, train the autoregressive model by using throughput data, and determine other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model;
a predicting module 1003, configured to predict, by using the intermediate model, an end of term balance of a specified date in the throughput data, to obtain a predicted end of term balance, where the specified date is one day of recording time in the throughput data;
a calculating module 1004, configured to calculate an evaluation index value of the intermediate model according to the throughput data and the predicted end-of-term balance;
a comparison module 1005, configured to compare the calculated evaluation index value with a preset standard index value range;
and the prediction module 1003 is configured to determine the intermediate model as a target autoregressive model when the evaluation index value is within the standard index value range, and predict the future single-day deposit estimate of the target date by using the target autoregressive model.
In an implementation manner of the embodiment of the present invention, the model training module 1002 is configured to:
setting corresponding parameters in the autoregressive model by using preset parameters;
dividing the throughput data into a training set and a testing set according to the sequence of the recording time according to a preset proportion, wherein the recording time of the throughput data in the training set is prior to the recording time of the throughput data in the testing set, and the preset proportion is used for indicating the proportion of the data volume in the testing set and the data volume in the training set to the total quantity of the throughput data respectively;
and training the autoregressive model with the preset parameters by using a training set, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model.
In one implementation of an embodiment of the present invention,
the predicting module 1003 is further configured to predict the end balance of the first specified date in the training set by using the intermediate model, so as to obtain a first predicted end balance;
a calculating module 1004, further configured to calculate a first evaluation index value of the intermediate model according to the training set and the first predicted end-of-term balance;
the comparison module 1005 is further configured to compare the calculated first evaluation index value with a preset standard index value range;
the prediction module 1003 is further configured to, if the first evaluation index value is within the standard index value range, predict the end balance of the second specified date in the test set by using the intermediate model to obtain a second predicted end balance;
a calculating module 1004, further configured to calculate a second evaluation index value of the intermediate model according to the test set and the second predicted end-of-term balance;
the comparison module 1005 is further configured to compare the calculated second evaluation index value with a preset standard index value range;
the predicting module 1003 is further configured to determine the intermediate model as a target autoregressive model when the second evaluation index value is within the standard index value range, and predict the deposit amount of the target date by using the target autoregressive model.
In an implementation manner of the embodiment of the present invention, the apparatus 1000 further includes:
the parameter adjusting module 1006 is configured to adjust the preset parameter when the evaluation index value is not within the standard index value range, reconstruct the autoregressive model by using the model training module 1002, and train the reconstructed autoregressive model by using throughput data until the evaluation index value determined according to the reconstructed autoregressive model is within the standard index value range.
In one implementation of an embodiment of the present invention,
an obtaining module 1001, configured to obtain throughput data of a target date when the target date is finished;
the obtaining module 1001 is further configured to add throughput data of a target date into the test set, select, according to the sequence of the recording time from first to last, throughput data of the same amount as the throughput data of the target date from the test set, remove the test set and move the test set into the training set, select, from the training set, throughput data of the same amount as the throughput data of the target date and move the training set out, and obtain an updated training set and an updated test set;
a calculating module 1004, configured to calculate an evaluation index value of the currently used target regression model by using the updated training set, the updated test set, and the predicted end-of-term balance of the target date predicted by the predicting module after acquiring the throughput data of the target date and updating the training set and the test set each time;
the prediction module 1003 is configured to continue to use the target autoregressive model when the evaluation index value of the currently used target autoregressive model is within the standard index value range, and trigger the acquisition module 1001 to update the training set and the test set every day until the time period for using the target autoregressive model reaches the maintenance date, adjust the preset parameters by the parameter adjusting module 1006, and re-determine the target autoregressive model by using the updated training set and the test set by the model training module 1003.
In one implementation of an embodiment of the present invention,
the parameter adjusting module 1006 is further configured to, when the evaluation index value of the currently used target regression model is not within the standard index value range, adjust the preset parameter, and trigger the model training module 1003 to re-determine the target autoregressive model by using the updated training set and the updated test set.
In the embodiment of the invention, the throughput data of the intelligent counter in the preset time period is obtained, the throughput data and the preset parameters are utilized to establish the autoregressive model, and an evaluation index value is calculated through the prediction result and the throughput data of the autoregressive model to evaluate the accuracy of the established autoregressive model, when the evaluation index value is determined to be in the standard index value range, the accuracy of the autoregressive model is determined to reach the standard, the method can predict the future single-day deposit forecast of the future fund of a certain target date, compared with the prediction by the experience of the staff in the prior art, the method can mine the periodic rule of the historical throughout data, therefore, the suitable spare payment money adding amount is found, the bank operation cost is reduced, the cash utilization efficiency of the intelligent counter is improved, the operation cost of the cash carrier is reduced, and the problem of low customer experience caused by the failure of money taking is avoided.
An embodiment of the present invention further provides a computer device, and fig. 11 is a schematic diagram of a computer device in an embodiment of the present invention, where the computer device is capable of implementing all steps in the spare payment management in the foregoing embodiment, and the computer device specifically includes the following contents:
a processor (processor)1101, a memory (memory)1102, a communication Interface (Communications Interface)1103, and a communication bus 1104;
the processor 1101, the memory 1102 and the communication interface 1103 complete mutual communication through the communication bus 1104; the communication interface 1103 is used for implementing information transmission between related devices;
the processor 1101 is configured to call a computer program in the memory 1102, and when the processor executes the computer program, the processor implements the reserve payment management in the above embodiments.
An embodiment of the present invention further provides a computer-readable storage medium, where a computer program for executing the method for managing a reserve payment is stored in the computer-readable storage medium.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (14)
1. A method for managing funds on deposit, the method comprising:
acquiring daily handling data of an intelligent counter of a bank outlet in a preset time period, wherein the handling data comprises deposit amount of each time of a deposit fund, withdrawal amount of each time, single-day end balance and corresponding recording time;
constructing an autoregressive model by using preset parameters, training the autoregressive model by using the throughput data, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model;
predicting the end balance of a specified date in the handling data by using an intermediate model to obtain a predicted end balance, wherein the specified date is one day of the recording time in the handling data;
calculating an evaluation index value of the intermediate model according to the throughput data and the predicted end balance;
comparing the calculated evaluation index value with a preset standard index value range;
and if the evaluation index value is in the standard index value range, determining the intermediate model as a target autoregressive model, and predicting the reserve payment single-day deposit estimate of the future target date by using the target autoregressive model.
2. The method of claim 1, wherein the steps of constructing an autoregressive model by using preset parameters, training the autoregressive model by using the throughput data, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model comprise:
setting corresponding parameters in the autoregressive model by using preset parameters;
dividing the throughput data into a training set and a testing set according to the sequence of the recording time according to a preset proportion, wherein the recording time of the throughput data in the training set is prior to the recording time of the throughput data in the testing set, and the preset proportion is used for indicating the proportion of the data volume in the testing set and the data volume in the training set in the total quantity of the throughput data respectively;
and training the autoregressive model with the preset parameters by using a training set, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model.
3. The method of claim 2, wherein after the auto-regression model with the preset parameters is trained by using a training set, and other parameters except the preset parameters in all undetermined parameters of the auto-regression model are determined to obtain an intermediate model, the method further comprises:
predicting the end balance of a first appointed date in the training set by using an intermediate model to obtain a first predicted end balance;
calculating a first evaluation index value of the intermediate model according to the training set and the first prediction end balance;
comparing the first evaluation index value obtained by calculation with a preset standard index value range;
if the first evaluation index value is within the standard index value range, predicting the end balance of a second specified date in the test set by using an intermediate model to obtain a second predicted end balance;
calculating a second evaluation index value of the intermediate model according to the test set and the second prediction end balance;
comparing the calculated second evaluation index value with a preset standard index value range;
and if the second evaluation index value is within the standard index value range, determining the intermediate model as a target autoregressive model, and predicting the deposit amount of the reserve payment on the target date by using the target autoregressive model.
4. The method according to any one of claims 1 to 3, further comprising:
and if the evaluation index value is not in the standard index value range, adjusting the preset parameters, reconstructing the autoregressive model, and training the reconstructed autoregressive model by using throughput data until the evaluation index value determined according to the reconstructed autoregressive model is in the standard index value range.
5. The method of any one of claims 2 or 3, wherein after predicting the payback gold single-day placement estimate for the target date using the target autoregressive model, the method further comprises:
when the target date is finished, acquiring throughput data of the target date;
adding the handling data of the target date into a test set, selecting the handling data with the same volume as the handling data of the target date from the test set according to the sequence of the recording time from first to last, removing the test set and moving the test set into a training set, selecting the handling data with the same volume as the handling data of the target date from the training set, and removing the training set to obtain an updated training set and a test set;
after acquiring the handling data of the target date and updating the training set and the test set each time, calculating the evaluation index value of the currently used target regression model by using the updated training set, the updated test set and the predicted end-of-term balance of the target date;
and if the evaluation index value of the currently used target regression model is within the standard index value range, continuing to use the target autoregressive model, updating the training set and the test set every day until the time length for using the target autoregressive model reaches the maintenance date, adjusting the preset parameters, and re-determining the target autoregressive model by using the updated training set and the test set.
6. The method of claim 5, wherein after calculating an evaluation index value for the currently used target regression model using the updated training set, test set, and predicted end of term balance for the predicted target date, the method further comprises:
and if the evaluation index value of the currently used target regression model is not in the standard index value range, adjusting the preset parameters, and re-determining the target autoregressive model by using the updated training set and the test set.
7. A pay-as-you-go management apparatus, the apparatus comprising:
the system comprises an acquisition module, a storage module and a management module, wherein the acquisition module is used for acquiring the daily handling data of an intelligent counter of a bank outlet in a preset time period, and the handling data comprises the deposit amount of each time of a deposit fund, the withdrawal amount of each time, the single-day end balance and the corresponding recording time;
the model training module is used for constructing an autoregressive model by using preset parameters, training the autoregressive model by using the throughput data, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model;
the forecasting module is used for forecasting the end balance of the appointed date in the handling data by utilizing the intermediate model to obtain the forecast end balance, wherein the appointed date is one day in the recording time in the handling data;
the calculation module is used for calculating an evaluation index value of the intermediate model according to the throughput data and the predicted end balance;
the comparison module is used for comparing the calculated evaluation index value with a preset standard index value range;
and the prediction module is used for determining the intermediate model as a target autoregressive model when the evaluation index value is within the standard index value range, and predicting the reserve fund single-day deposit estimate of the future target date by using the target autoregressive model.
8. The apparatus of claim 7, wherein the model training module is configured to:
setting corresponding parameters in the autoregressive model by using preset parameters;
dividing the throughput data into a training set and a testing set according to the sequence of the recording time according to a preset proportion, wherein the recording time of the throughput data in the training set is prior to the recording time of the throughput data in the testing set, and the preset proportion is used for indicating the proportion of the data volume in the testing set and the data volume in the training set in the total quantity of the throughput data respectively;
and training the autoregressive model with the preset parameters by using a training set, and determining other parameters except the preset parameters in all undetermined parameters of the autoregressive model to obtain an intermediate model.
9. The apparatus of claim 8,
the prediction module is further used for predicting the end balance of the first appointed date in the training set by using the intermediate model to obtain a first predicted end balance;
the calculation module is also used for calculating a first evaluation index value of the intermediate model according to the training set and the first prediction end balance;
the comparison module is also used for comparing the first evaluation index value obtained by calculation with a preset standard index value range;
the prediction module is further used for predicting the end balance of a second specified date in the test set by using the intermediate model to obtain a second predicted end balance if the first evaluation index value is in the standard index value range;
the calculation module is also used for calculating a second evaluation index value of the intermediate model according to the test set and the second prediction end balance;
the comparison module is also used for comparing the calculated second evaluation index value with a preset standard index value range;
and the prediction module is further used for determining the intermediate model as a target autoregressive model when the second evaluation index value is within the standard index value range, and predicting the deposit amount of the reserve payment on the target date by using the target autoregressive model.
10. The apparatus of any one of claims 7 to 9, further comprising:
and the parameter adjusting module is used for adjusting the preset parameters when the evaluation index value is not in the standard index value range, reconstructing the autoregressive model by using the model training module, and training the reconstructed autoregressive model by using the throughput data until the evaluation index value determined according to the reconstructed autoregressive model is in the standard index value range.
11. The apparatus according to any one of claims 7 or 8,
the acquisition module is also used for acquiring the handling data of the target date when the target date is finished;
the acquisition module is also used for adding the handling data of the target date into the test set, selecting the handling data with the same volume as the handling data of the target date from the test set according to the sequence of the recording time from first to last, removing the test set and moving the test set into the training set, selecting the handling data with the same volume as the handling data of the target date from the training set and moving the training set out to obtain an updated training set and an updated test set;
the calculation module is used for calculating the evaluation index value of the currently used target regression model by using the updated training set, the updated test set and the predicted end balance of the target date predicted by the prediction module after acquiring the handling data of the target date and updating the training set and the test set each time;
and the prediction module is used for continuously using the target autoregressive model when the evaluation index value of the currently used target autoregressive model is within the standard index value range, triggering the acquisition module to update the training set and the test set every day until the time for using the target autoregressive model reaches the maintenance date, adjusting the preset parameters by the parameter adjusting module, and re-determining the target autoregressive model by using the updated training set and the test set by the model training module.
12. The apparatus of claim 11,
and the parameter adjusting module is also used for adjusting the preset parameters and triggering the model training module to re-determine the target autoregressive model by using the updated training set and the test set when the evaluation index value of the currently used target regression model is not within the standard index value range.
13. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 6 when executing the computer program.
14. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110856998.5A CN113487110A (en) | 2021-07-28 | 2021-07-28 | Spare payment management method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110856998.5A CN113487110A (en) | 2021-07-28 | 2021-07-28 | Spare payment management method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113487110A true CN113487110A (en) | 2021-10-08 |
Family
ID=77943120
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110856998.5A Pending CN113487110A (en) | 2021-07-28 | 2021-07-28 | Spare payment management method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113487110A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115662143A (en) * | 2022-11-21 | 2023-01-31 | 吉林大学 | Dynamic prediction system and method for operation safety situation of public transport enterprise |
-
2021
- 2021-07-28 CN CN202110856998.5A patent/CN113487110A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115662143A (en) * | 2022-11-21 | 2023-01-31 | 吉林大学 | Dynamic prediction system and method for operation safety situation of public transport enterprise |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104424598A (en) | Cash demand quantity predicating device and method | |
| CN111382911B (en) | High-rise personnel scheduling prediction method based on banking outlet business data | |
| CN109544316B (en) | Method and system for urging collection of real-time case division according to proportion | |
| CN114118691A (en) | Scheduling scheme processing method, device, equipment and medium based on traffic | |
| CN107491563A (en) | Towards the data processing method and system of settlement for account | |
| US8219368B1 (en) | Capacity modeling system | |
| CN108282586B (en) | Prediction technique, device, storage medium and terminal based on service line | |
| CN113487110A (en) | Spare payment management method and device | |
| CN107977855A (en) | A kind of method and device of managing user information | |
| US20220398604A1 (en) | Systems and methods for dynamic cash flow modeling | |
| CN110532156A (en) | A kind of capacity prediction methods and device | |
| CN109543865A (en) | Reserve transaction processing method and device | |
| CN111738507A (en) | Bank clearing position fund payment amount prediction method, device, equipment and medium | |
| CN112862197A (en) | Intelligent network point number allocation method, device, equipment and storage medium | |
| Mikosch et al. | Real-time forecasting with a MIDAS VAR | |
| CN112037049B (en) | Bank card charging method and device | |
| CN107274143A (en) | Settlement processing method and device | |
| CN113610323A (en) | Cash demand prediction method and device for bank self-service equipment | |
| CN111861544A (en) | Participant account liquidity prediction method and device | |
| CN114900443B (en) | Method and device for establishing incoming line traffic prediction model | |
| CN117236804B (en) | Financial enterprise performance identification method, equipment and medium | |
| CN113361829B (en) | Time adjustment method for contract energy management based on artificial intelligence and big data | |
| Kundu et al. | Forecasting economic indicators of Bangladesh using vector autoregressive (VAR) model | |
| CN113988461A (en) | Position prediction method, position prediction device, storage medium and electronic equipment | |
| Uzor | 4 African small and medium enterprises and the challenges in global value chains |
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 |