CN115393037A

CN115393037A - Bank outlet cash allocation method and device

Info

Publication number: CN115393037A
Application number: CN202211016744.3A
Authority: CN
Inventors: 龚孟旭
Original assignee: Bank of China Ltd
Current assignee: Bank of China Ltd
Priority date: 2022-08-24
Filing date: 2022-08-24
Publication date: 2022-11-25

Abstract

The invention provides a cash allocation method and a cash allocation device for bank outlets, and particularly relates to the field of finance, wherein the method comprises the following steps: obtaining a training sample according to the historical operation characteristic vectors of the target bank outlets in a plurality of periods; training a preset sequence processing model by using the training sample to obtain a mature model; obtaining a predicted cash demand interval of the next period of the target bank outlet according to the current operation characteristic vectors of the target bank outlet in a plurality of periods and the mature model; and judging whether the cash reserve volume of the target bank outlet is in the predicted cash demand volume interval, if not, feeding back corresponding warning information to the staff based on the cash reserve volume and the predicted cash demand volume interval so that the staff can allocate the cash of the target bank outlet. The invention can improve the speed and the accuracy of cash allocation, thereby improving the efficiency of cash allocation and further improving the operation efficiency of banks and the experience of users.

Description

Bank outlet cash allocation method and device

Technical Field

The invention relates to the technical field of money distribution scheduling, in particular to the field of finance, and particularly relates to a cash distribution method and device for a bank outlet.

Background

When a bank outlet meets the requirement of a large number of users for withdrawing cash, in order to enable the cash reserve to meet the corresponding requirement as much as possible, cash allocation is often performed in advance. In the prior art, the cash allocation mainly includes that related staff analyze historical operation conditions of banks, future cash demand is predicted and determined, and cash is scheduled based on the predicted cash demand and current cash reserve conditions. However, the manual analysis and prediction needs to take a long time, which leads to a tight time for scheduling cash in the following process, and is not favorable for meeting the corresponding withdrawal requirement. Moreover, the accuracy of the manual analysis depends mainly on the working experience and the working ability of the staff, so that the accuracy of the cash demand predicted by the analysis is low, and the cash is probably not enough to meet the corresponding cash withdrawal requirement in scheduling. In summary, in the prior art, the cash allocating speed is low, and the accuracy is low, so that the cash allocating efficiency is low, and the operation efficiency of the bank and the experience of the user are not improved.

Disclosure of Invention

The invention aims to provide a cash allocating method for bank outlets, which aims to solve the problems that the existing cash allocating method is low in speed and accuracy, so that the cash allocating efficiency is low, and the operation efficiency of banks and the experience of users are not improved. Another object of the present invention is to provide a cash dispenser for banking outlets. It is a further object of this invention to provide such a computer apparatus. It is a further object of this invention to provide such a readable medium. It is a further object of this invention to provide a computer program product.

In order to achieve the above object, an aspect of the present invention discloses a cash allocating method at a banking outlet, where the method includes:

obtaining a training sample according to the historical operation characteristic vectors of the target bank outlets in a plurality of periods;

training a preset sequence processing model by using the training sample to obtain a mature model; obtaining a predicted cash demand interval of the next period of the target bank outlet according to the current operation characteristic vectors of the target bank outlet in a plurality of periods and the mature model;

and judging whether the cash reserve volume of the target bank outlet is in the predicted cash demand volume interval, if not, feeding back corresponding warning information to the staff based on the cash reserve volume and the predicted cash demand volume interval so that the staff can allocate the cash of the target bank outlet.

Optionally, further comprising:

before obtaining a training sample according to the historical operation characteristic vectors of a plurality of periods of a target bank branch,

and performing feature extraction processing on the historical operation information of the target bank outlet in multiple periods to obtain corresponding historical operation feature vectors.

Optionally, further comprising:

before the forecast cash demand interval of the next period of the target bank branch is obtained according to the current operation characteristic vector of the target bank branch in a plurality of periods and the mature model,

and performing feature extraction processing on the current operation information of the target bank outlet in multiple periods to obtain the corresponding current operation feature vector.

Optionally, the obtaining a training sample according to the historical operation feature vectors of the target banking outlet in multiple cycles includes:

taking historical operation characteristic vectors of a target bank outlet in multiple periods as first dimension vectors of a two-dimensional matrix, and obtaining a current historical operation matrix according to the first dimension vectors;

repeatedly executing a matrix shifting step of a preset shifting number of times, wherein the matrix shifting step comprises the following steps: shifting the current historical operation matrix by a preset unit along a second dimension direction to obtain a dislocation matrix, and taking the dislocation matrix as the current historical operation matrix;

splicing the current historical operation matrix and the plurality of dislocation matrixes along a first dimension direction to obtain a historical data matrix;

and obtaining a training sample based on the first dimension vector of the historical data matrix.

Optionally, the splicing the current historical operation matrix and the plurality of dislocation matrices along the first dimension direction to obtain the historical data matrix includes:

and splicing the current historical operation matrix and the plurality of dislocation matrices along the direction of the first dimension according to the time sequence of the corresponding period of the corresponding first dimension vector to obtain the historical data matrix.

Optionally, obtaining a training sample based on the first dimension vector of the historical data matrix includes:

respectively obtaining corresponding historical amount demand intervals based on a plurality of first dimension vectors of the historical data matrix;

taking the first dimension vector of the historical data matrix as an input sample, and taking the corresponding historical amount demand interval as a corresponding output sample;

and obtaining corresponding training samples based on a plurality of input samples and corresponding output samples respectively.

Optionally, the obtaining the corresponding historical amount demand intervals based on the plurality of first dimension vectors of the historical data matrix respectively includes:

obtaining the latest operation cycle corresponding to the first dimension vector of the historical data matrix based on the first dimension vector of the historical data matrix;

and obtaining the corresponding historical amount demand interval based on the actual amount demand of the next cycle of the latest operation cycle.

Optionally, the obtaining a predicted cash demand interval of a next period of the target banking outlet according to the current operation feature vector of the target banking outlet in multiple periods and the maturity model includes:

obtaining a current input vector according to the current operation characteristic vector and the time of the corresponding period;

and inputting the current input vector into the maturity model so that the maturity model is operated to obtain the predicted cash demand interval.

Optionally, the feeding back corresponding warning information to the staff based on the cash reserve volume and the predicted cash demand volume interval to enable the staff to allocate the cash of the target bank outlet includes:

judging whether the cash reserve amount is smaller than the lower limit of the predicted cash demand interval or not;

if yes, feeding back insufficient cash reserve warning information to the staff according to the cash reserve amount and the predicted cash demand interval, so that the staff allocate cash based on the insufficient cash reserve warning information.

Optionally, further comprising:

when the cash reserve amount is judged to be not less than the lower limit of the predicted cash demand interval, judging whether the cash reserve amount is greater than the upper limit of the predicted cash demand interval;

and if so, feeding back cash reserve excess alarm information to the staff according to the cash reserve amount and the predicted cash demand interval so that the staff allocate cash based on the cash reserve excess alarm information.

In order to achieve the above object, another aspect of the present invention discloses a cash allocating apparatus for a banking outlet, the apparatus comprising:

the training sample generation module is used for obtaining training samples according to the historical operation characteristic vectors of the target bank outlets in a plurality of periods;

the prediction module is used for training a preset sequence processing model by using the training sample to obtain a mature model; obtaining a predicted cash demand interval of the next period of the target bank outlet according to the current operation characteristic vectors of the target bank outlet in a plurality of periods and the mature model;

and the allocation module is used for judging whether the cash reserve volume of the target bank outlet is in the predicted cash demand volume interval or not, and if not, feeding back corresponding alarm information to the staff based on the cash reserve volume and the predicted cash demand volume interval so that the staff allocate cash of the target bank outlet.

The invention also discloses 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 when executing the program.

The invention also discloses a computer-readable medium, on which a computer program is stored which, when executed by a processor, implements a method as described above.

The invention also discloses a computer program product comprising a computer program which, when executed by a processor, implements the method as described above.

According to the bank branch cash allocation method and device, the training sample is obtained according to the historical operation characteristic vectors of the target bank branch in a plurality of periods, the training sample can be obtained by taking the historical operation characteristic vectors which fully reflect the past practical operation conditions of the bank branch as input, and the obtained training sample has high authenticity, so that the prediction accuracy of the mature model obtained by training in the subsequent steps can be improved, and the accuracy of the whole cash allocation can be improved. Training a preset sequence processing model by using the training sample to obtain a mature model; according to the current operation characteristic vectors of the target bank branch in multiple cycles and the mature model, the predicted cash demand interval of the next cycle of the target bank branch is obtained, accurate training samples with high authenticity can be used for training the model, the model has high prediction accuracy and calculation speed, the speed and the accuracy of the predicted cash demand interval are greatly improved by means of the advantages of high automation degree, high calculation speed and high calculation accuracy of the model, and the model is a sequence processing model suitable for processing the characteristic vectors, so that the speed and the accuracy of the predicted cash demand interval can be further improved, and the speed and the accuracy of cash allocation are improved. Whether the cash reserve volume of the target bank outlet is in the predicted cash demand volume interval or not is judged, if not, corresponding warning information is fed back to a worker based on the cash reserve volume and the predicted cash demand volume interval, so that the worker can allocate cash of the target bank outlet, the worker can be automatically, quickly and accurately guided to allocate the cash based on the predicted cash demand volume interval and the actual cash reserve volume, the worker does not need to analyze and predict the corresponding cash demand volume in person, and the speed and the accuracy of overall cash allocation are improved. In summary, the method and the device for allocating cash at a bank outlet provided by the invention can improve the speed and accuracy of allocating cash, thereby improving the efficiency of allocating cash, and further improving the operating efficiency of a bank and the experience of a user.

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.

Fig. 1 is a schematic flow chart illustrating a cash allocating method of a banking outlet according to an embodiment of the present invention;

FIG. 2 is a schematic diagram illustrating an alternative procedure for obtaining training samples according to an embodiment of the present invention;

FIG. 3 is a diagram illustrating an alternative step of obtaining a historical monetary demand interval, in accordance with an embodiment of the present invention;

FIG. 4 is a diagram illustrating an alternative step of obtaining a predicted cash demand interval according to an embodiment of the present invention;

FIG. 5 is a diagram illustrating an optional step of feeding back alarm information according to an embodiment of the present invention;

FIG. 6 is a block diagram of a cash dispenser at a banking outlet according to an embodiment of the present invention;

FIG. 7 illustrates a schematic diagram of a computer device suitable for use in implementing embodiments of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The terms "first," "second," "8230," "8230," and the like, as used herein, are not intended to be limited to a specific meaning or sequence, nor are they intended to limit the invention, but only to distinguish one element from another or to distinguish one element from another element.

As used herein, the terms "comprising," "including," "having," "containing," and the like are open-ended terms that mean including, but not limited to.

As used herein, "and/or" includes any and all combinations of the described items.

It should be noted that, in the technical solution of the present invention, the acquisition, storage, use, processing, etc. of the data all conform to the relevant regulations of the national laws and regulations.

The embodiment of the invention discloses a cash allocation method for bank outlets, which specifically comprises the following steps of:

s101: and obtaining a training sample according to the historical operation characteristic vectors of the target bank outlets in a plurality of periods.

S102: training a preset sequence processing model by using the training sample to obtain a mature model; and obtaining a predicted cash demand interval of the next period of the target bank branch according to the current operation characteristic vectors of the target bank branch in a plurality of periods and the mature model.

S103: and judging whether the cash reserve volume of the target bank outlet is in the predicted cash demand volume interval, if not, feeding corresponding warning information back to the staff based on the cash reserve volume and the predicted cash demand volume interval so that the staff allocates cash of the target bank outlet.

For example, the first dimension and the second dimension in the embodiment of the present invention may be, but are not limited to, a row or a column. Wherein the first dimension is different from the second dimension. For example, when the first dimension is a row, the second dimension is a column, and correspondingly, the first dimension vector refers to a row vector, the second dimension vector refers to a column vector, the first dimension direction refers to a row direction (which can be understood as, but not limited to, a horizontal direction in a matrix), and the second dimension direction refers to a column direction (which can be understood as, but not limited to, a vertical direction in the matrix). For another example, when the first dimension is a column, the second dimension is a row, and correspondingly, the first dimension vector refers to a column vector, the second dimension vector refers to a row vector, the first dimension direction refers to a column direction (which can be understood as but not limited to a vertical direction in a matrix), and the second dimension direction refers to a row direction (which can be understood as but not limited to a horizontal direction in the matrix). It should be noted that the specific references to the first dimension and the second dimension can be determined by those skilled in the art according to practical situations, and the above description is only an example and is not limiting.

Illustratively, the sequence processing model may be, but is not limited to, a recurrent neural network model (RNN model), a long-short term memory model (LSTM model), or the like, and is preferably a long-short term memory model (LSTM model). It should be noted that the kind of the sequence processing model can be determined by those skilled in the art according to actual situations, and the above description is only an example and is not limited thereto.

Illustratively, training the model with training samples is a conventional technique in the art and will not be described here.

Illustratively, the period may be, but is not limited to, a week, month, day, etc., preferably a week. Each period corresponds to a time, and the time is a certain time in the period, for example, a certain period is a certain week of 2021 year 3 month, and the range is from 11 days of 2021 year 3 month to 17 days of 2021 year 3 month, then the time corresponding to the period may be, but is not limited to, 11 days of 2021 year 3 month or 17 days of 2021 year 3 month, etc. It should be noted that, the nature of the period and the time corresponding to the period can be determined by those skilled in the art according to actual situations, and the above description is only an example, and is not limited thereto.

For example, the type of cash in the embodiment of the present invention may be, but is not limited to, domestic currency or foreign currency of different types. The type of cash may be determined by those skilled in the art according to actual circumstances, and the above description is only an example and is not intended to limit the scope of the present invention.

In an optional embodiment, further comprising:

and performing feature extraction processing on the historical operation information of the target bank outlets in multiple cycles to obtain corresponding historical operation feature vectors.

Illustratively, one period corresponds to one piece of historical operation information, and one period corresponds to one historical operation feature vector.

Illustratively, the business information includes, but is not limited to, the total number of users in a single period of the target banking site, the average savings amount in a single period, the population in the site coverage (which may be obtained by, but is not limited to, multiplying a preset site coverage area by a corresponding population density, wherein the population density may be determined by surveying visits or querying related population information on related websites), the number of communities in the site coverage, the actual cash demand in a single period, and whether the period is in a holiday. It should be noted that, the specific content of the operation information can be determined by those skilled in the art according to the actual situation, and the above description is only an example, and is not limited thereto.

For example, the feature extraction processing is performed on the historical operation information of the target bank outlet in multiple cycles to obtain the corresponding historical operation feature vector, which may be, but is not limited to, performing corresponding format adjustment and feature extraction on each attribute information in the historical operation information to obtain vector elements corresponding to the attribute information, and then splicing and integrating each vector element to obtain the corresponding historical operation feature vector. For attribute information of which the attribute value is a number in the attribute information, when determining the vector element, the attribute value serving as the number may be directly determined as the corresponding vector element, or the attribute value is normalized to obtain the corresponding vector element; for attribute information whose attribute value is not a number in the attribute information (e.g., attribute value of a character, or an expression type), when determining a vector element, the attribute information may be digitized to obtain a corresponding vectorized element, and the digitized product may be normalized to obtain a corresponding vectorized element, and the implementation of the digitization may be, but not limited to, processing using a digitization coding (e.g., one-hot) algorithm or comparing to an ASCII code table. It should be noted that, the specific implementation manner of obtaining the corresponding historical operation feature vector by performing feature extraction processing on the historical operation information of the target bank branch in multiple cycles may be determined by those skilled in the art according to actual situations, and the above description is only an example, and does not limit the present invention.

Preferably, before the historical operation information of a plurality of cycles of the target bank outlet is subjected to the feature extraction processing, data preprocessing operation such as data cleaning can be further carried out on the historical operation information so as to eliminate redundant information in the historical operation information and correct wrong information.

Through the steps, the actual historical operation information can be converted into a vector form which is convenient to participate in calculation and processing, so that the difficulty of related calculation and processing operations in subsequent steps is reduced, the accuracy of the calculation and processing operations is improved, and the speed and the accuracy of overall cash allocation can be improved.

In an optional embodiment, further comprising:

Illustratively, one period corresponds to one piece of current operation information, and one period corresponds to one current operation feature vector.

Illustratively, the business information includes, but is not limited to, the total number of users in a single period of a target bank outlet, the average savings amount in a single period, the population in the outlet coverage (which may be obtained by, but is not limited to, multiplying a preset area of the outlet coverage by a corresponding population density, wherein the population density may be determined by a survey visit or by querying related information about population quantity on related websites), the number of communities in the outlet coverage, the actual cash demand in a single period, and whether the period is in a holiday. It should be noted that, the specific content of the operation information can be determined by those skilled in the art according to the actual situation, and the above description is only an example, and is not limited thereto.

For example, the feature extraction processing is performed on the current operation information of the target bank outlet in multiple cycles to obtain the corresponding current operation feature vector, which may be, but is not limited to, performing corresponding format adjustment and feature extraction on each attribute information in the current operation information to obtain vector elements corresponding to the attribute information, and then splicing and integrating each vector element to obtain the corresponding current operation feature vector. For attribute information of which the attribute value is a number in the attribute information, when determining the vector element, the attribute value serving as the number may be directly determined as the corresponding vector element, or the attribute value is normalized to obtain the corresponding vector element; for attribute information whose attribute value is not a number in the attribute information (e.g., attribute value of a word, a character, or an expression type), when determining a vector element, the attribute information may be digitized to obtain a corresponding vectorized element, and the digitized product may be normalized to obtain a corresponding vectorized element, where the implementation of the digitizing may be, but not limited to, processing using a digital coding (e.g., one-hot) algorithm or comparing to an ASCII code table. It should be noted that, the specific implementation manner of obtaining the corresponding current operation feature vector by performing feature extraction processing on the current operation information of the target bank branch in multiple cycles may be determined by those skilled in the art according to actual situations, and the above description is only an example, and does not limit this.

Preferably, before the current operation information of the target bank outlet in multiple cycles is subjected to the feature extraction processing, a data preprocessing operation such as data cleaning can be further performed on the current operation information, so that redundant information in the current operation information is eliminated and wrong information is corrected.

Through the steps, the actual current operation information can be converted into a vector form which is convenient to participate in calculation and processing, so that the difficulty of related calculation and processing operations in subsequent steps is reduced, the accuracy of the calculation and processing operations is improved, and the speed and the accuracy of the whole cash allocation can be improved.

In an optional embodiment, the obtaining a training sample according to the historical operating feature vector of the target banking site for a plurality of periods includes:

repeatedly executing a matrix shifting step for a preset number of shifts, the matrix shifting step comprising: shifting the current historical operation matrix by a preset unit along a second dimension direction to obtain a dislocation matrix, and taking the dislocation matrix as the current historical operation matrix;

Illustratively, the historical operation characteristic vectors of a plurality of periods of the target bank outlet are used as first dimension vectors of a two-dimensional matrix, and the current historical operation matrix is obtained according to the plurality of first dimension vectors, which may be but is not limited to using the historical operation characteristic vectors as the first dimension vectors, and the current operation matrix is obtained by splicing and integrating along a second dimension direction according to a sequence of time corresponding to the period to which the historical operation characteristic vectors belong (which may be but is not limited to a sequence from morning to evening or a sequence from evening to morning, etc.). The specific number of the plurality of cycles may be determined by those skilled in the art according to practical situations, and the embodiment of the present invention is not limited thereto, for example, the specific number of the plurality of cycles may be, but is not limited to, 52, 26, 13, 4, or 104, and is preferably 52. Taking the historical operation characteristic vectors of a plurality of periods of a target bank outlet as first dimension vectors of a two-dimensional matrix, and obtaining a current historical operation matrix according to the first dimension vectors, wherein the method comprises the following steps:

the plurality of historical operation feature vectors are specifically historical operation feature vectors corresponding to 52 weeks (52 weeks) in 2021, and the first dimension is set as rows and the second dimension is set as columns. Wherein, the historical operation feature vector corresponding to the 1 st week is V1 (1001, 10000000, 50000, 100), the historical operation feature vector corresponding to the 2 nd week is V2 (1100, 12000000, 45000, 90), the historical operation feature vector corresponding to the 3 rd week is V3 (900, 11000000, 46000, 85), the historical operation feature vector corresponding to the 4 th week is V4 (950, 9780000, 40000, 70) \8230, the historical operation feature vector corresponding to the 52 th week is V52 (1200, 13100000, 42500, 110), the vectors V1 to V52 are all adjusted to be in the form of row vectors to be used as row vectors of a two-dimensional matrix, and then the current historical operation matrix is obtained by splicing:

the current historical business matrix can also be simplified as:

here, the historical operation feature vectors of weeks 5 to 51 are not shown in the embodiment of the present invention.

It should be noted that, for a specific implementation manner that the historical operation feature vectors of multiple cycles of the target bank outlet are used as the first dimension vectors of the two-dimensional matrix and the current historical operation matrix is obtained according to multiple first dimension vectors, the implementation manner may be determined by a person skilled in the art according to actual situations, and the foregoing description is only an example, and does not limit the description.

For example, if the second dimension is a row, the second dimension direction is a row direction, and may be, but is not limited to, a left direction or a right direction, and the like, and if the second dimension is a column, the second dimension direction is a column direction, and may be, but is not limited to, an upward direction or a downward direction, and the like. It should be noted that, the second dimension direction can be determined by those skilled in the art according to practical situations, and the above description is only an example and is not limiting.

For example, the preset unit may be, but is not limited to, 1 row (or column), 2 rows (or columns), or 3 rows (or columns), and is preferably 1 row.

Illustratively, the preset number of shifts is equal to a multiple of a length of the first dimension vector of the expected subsequent historical data matrix (the expected subsequent single input to the model) divided by a length of the first dimension vector of the current historical business matrix (indicating that the first dimension vector of the historical data matrix includes several cycles of the historical business feature vectors) minus 1, for example, if the length of the first dimension vector of the expected subsequent historical data matrix is 4 times of the length of the first dimension vector of the current historical business matrix (indicating that the first dimension vector of the historical data matrix includes 4 cycles of the historical business feature vectors), the preset number of shifts is 3. It should be noted that the specific value of the preset shift number can be determined by those skilled in the art according to practical situations, and the above description is only an example and is not limited thereto.

For example, the specific manner of obtaining the misalignment matrix by the shift presetting unit may be, but is not limited to, a complementary code (zero padding) shift, a cyclic shift, or the like.

Illustratively, the matrix shifting step is repeatedly performed for a predetermined number of shifts, as follows:

given that the preset shift times is 3, the second dimension direction is the column direction (upward), the preset unit is 1 row, the shift mode is complement shift, and the initial current historical operation matrix is:

shifting the current historical operation matrix along a second dimension direction by a preset unit for the 1 st time to obtain a dislocation matrix as follows:

taking the dislocation matrix as the current historical operation matrix, shifting the current historical operation matrix along the second dimension direction for the 2 nd time by a preset unit to obtain the dislocation matrix as follows:

taking the dislocation matrix as the current historical operation matrix, shifting the current historical operation matrix along the second dimension direction for the 3 rd time by a preset unit to obtain a dislocation matrix which is:

accumulating the current historical operation matrix and the plurality of dislocation matrices, including:

for a total of 4 matrices.

It should be noted that, for a specific implementation manner of the matrix shift step repeatedly performing the preset shift times, the implementation manner can be determined by those skilled in the art according to practical situations, and the above description is only an example, and does not limit this.

Through the steps, the historical operation characteristic vectors in multiple periods can be integrated more orderly and more quickly in a matrix shifting and splicing mode, each first-dimension vector of the obtained historical data matrix comprises the historical operation characteristic vectors in the multiple periods which are different from each other, and the quantity of the historical operation characteristic vectors of each first-dimension vector is the same, so that when the training sample is determined subsequently, the different first-dimension vectors in the historical data matrix can be directly extracted to obtain partial samples, the samples do not need to be combed, searched and sorted for a long time, and the speed of determining the training sample is improved. And the extracted partial samples comprise different historical operation characteristic vectors of multiple periods with expected input quantity (the quantity meets the training requirement and the prediction requirement so as to enable the subsequent prediction accuracy to meet the requirement), so that the partial samples can better meet the subsequent training requirement, and the efficiency of the subsequent training model is improved so as to improve the operation speed and the accuracy of the trained model. Therefore, the above steps can improve the speed and accuracy of the overall cash allocation.

In an optional embodiment, the splicing the current historical operation matrix and the plurality of misalignment matrices along the first dimension direction to obtain a historical data matrix includes:

For example, if the first dimension is a row, the first dimension direction is a row direction, and may be, but is not limited to, a left direction or a right direction, and the like, and if the first dimension is a column, the first dimension direction is a column direction, and may be, but is not limited to, an upward direction or a downward direction, and the like. It should be noted that, the first dimension direction can be determined by those skilled in the art according to practical situations, and the above description is only an example and is not limiting.

For example, the sequence of the time of the corresponding period may be, but is not limited to, from morning to evening or from evening to morning.

Exemplarily, a plurality of matrices are spliced to obtain a large matrix, which is a conventional technical means in the art and is not described herein again.

Illustratively, the first dimension vector may be specifically interpreted as: the first dimension vector is a first row vector in the matrix (i.e., a row vector positioned at the top of the matrix) when the first dimension is a row, and the first dimension vector is a first column vector in the matrix (i.e., a column vector positioned at the left most position in the matrix) when the first dimension is a column. It should be noted that the specific meaning of the first dimension vector can be determined by those skilled in the art according to practical situations, and the above description is only an example and is not limiting.

For example, the current historical operation matrix and the plurality of misalignment matrices are spliced along the first dimension direction according to the time sequence of the corresponding period of the corresponding first dimension vector, so as to obtain the historical data matrix, where the historical data matrix has the following examples:

the first dimension direction is known as the right (row direction). The first dimension is a row.

The current historical operation matrix and a plurality of dislocation matrices are accumulated to include:

the total number of the 4 matrixes is 4, the first dimension vector of the matrix VT1 is V1, the first dimension vector of the matrix VT2 is V2, the first dimension vector of the matrix VT3 is V3, the first dimension vector of the matrix VT4 is V4, and the V1, V2, V3 and V4 are sorted into V1, V2, V3 and V4 according to the sequence (from morning to evening) of the time of the corresponding cycle, and then the splicing sequence is determined to be VT1, VT2, VT3 and VT4. The historical data matrix obtained after splicing is specifically as follows:

the historical data matrix can also be simplified as:

(VT1VT2VT3VT4)

it should be noted that, for the specific implementation manner of obtaining the historical data matrix by splicing the current historical operation matrix and the plurality of misalignment matrices along the first dimension direction according to the time sequence of the corresponding period of the corresponding first dimension vector, the specific implementation manner may be determined by those skilled in the art according to the actual situation, and the above description is only an example, and does not limit this.

Through the steps, each first dimension vector of the historical data matrix comprises a plurality of periods of historical operation characteristic vectors which are different from each other, and the plurality of periods of historical operation characteristic vectors are arranged according to the corresponding time sequence in each first dimension vector of the historical data matrix, so that the complexity of related sorting operation in the subsequent model training process is reduced, the speed and the accuracy of model training can be improved through the steps, and the speed and the accuracy of whole cash allocation are improved.

In an alternative embodiment, as shown in fig. 2, the obtaining a training sample based on the first dimension vector of the historical data matrix includes the following steps:

s201: and respectively obtaining corresponding historical money demand intervals based on the plurality of first dimension vectors of the historical data matrix.

S202: and taking the first dimension vector of the historical data matrix as an input sample, and taking the corresponding historical amount demand interval as a corresponding output sample.

S203: and obtaining corresponding training samples based on a plurality of input samples and corresponding output samples respectively.

Illustratively, one input sample corresponds to one output sample, i.e., one first dimension vector in the historical data matrix corresponds to one historical monetary demand interval.

For example, the obtaining of the corresponding training samples based on the plurality of input samples and the corresponding output samples, respectively, may be, but is not limited to, integrating the plurality of input samples and the corresponding output samples to obtain the corresponding training samples. One training sample may include an input sample and an output sample corresponding to the input sample. It should be noted that, for a specific implementation manner of obtaining corresponding training samples based on a plurality of input samples and corresponding output samples, respectively, may be determined by those skilled in the art according to actual situations, and the above description is only an example, and does not limit this.

Through the steps, the first dimension vector in the historical data matrix is obtained, and one first dimension vector in the historical data matrix corresponds to the historical operation feature vectors of a plurality of cycles, so that the determined historical amount demand interval can accurately correspond to the historical operation feature vectors of the plurality of cycles, and the subsequent training sample formed by taking the historical amount demand interval as an output sample can accurately meet the law and logic for predicting the cash demand, namely, the training process can accurately simulate the function of predicting the cash demand of a certain future cycle by using the known operation feature vectors of the plurality of cycles, and the cash demand comprises the cash demand (the cash demand is greater than the cash demand), so that the cash demand interval predicted by using the historical amount demand interval as a standard output training model can be larger during actual use, so that when the subsequent cash is dispatched, cash can be dispatched to a target bank network node for more cash, and the cash reserve of the target bank network node is reserved to deal with the situation that additional needs to be reserved (for example, the cash demand of another user is not reserved, and the cash demand cannot be reserved temporarily. Therefore, the accuracy of obtaining the corresponding training samples can be improved through the steps, the prediction speed and the prediction accuracy of the model obtained through training are improved, the actual situation is fully considered, the improvement of the user experience is facilitated, and the speed and the accuracy of the whole cash allocation are improved.

In an optional embodiment, as shown in fig. 3, the obtaining the corresponding historical amount demand intervals based on the plurality of first dimension vectors of the historical data matrix respectively includes the following steps:

s301: and obtaining the latest operation cycle corresponding to the first dimension vector of the historical data matrix based on the first dimension vector of the historical data matrix.

S302: and obtaining the corresponding historical amount demand interval based on the actual amount demand of the next cycle of the latest operation cycle.

For example, the latest operation cycle corresponding to the first dimension vector of the historical data matrix is obtained based on the first dimension vector of the historical data matrix, which may be, but is not limited to, determining a plurality of cycle historical operation eigenvectors included in each first dimension vector of the historical data matrix based on a plurality of first dimension vectors of the historical data matrix, where since the cycle to which the historical operation eigenvector belongs is known and the time corresponding to the cycle to which the historical operation eigenvector belongs is known, the time corresponding to a single historical operation eigenvector is also known, and the cycle to which the historical operation eigenvector corresponding to the latest time corresponds may be determined as the latest operation cycle of the corresponding first dimension vector of the historical data matrix. It should be noted that, for a specific implementation manner of obtaining the latest business cycle corresponding to the first dimension vector of the historical data matrix based on the first dimension vector of the historical data matrix, the specific implementation manner may be determined by a person skilled in the art according to an actual situation, and the above description is only an example, and does not limit this.

For example, if the latest period of operation is the 1 st week of the 2021 year 3 month (assuming the period range is 2021 year 3 month 1 day to 2021 year 3 month 7 days), then the corresponding next period is the 2 nd week of the 2021 year 3 month (correspondingly, the period range is 2021 year 3 month 8 days to 2021 year 3 month 14 days). It should be noted that the specific meaning of the next cycle of the latest business cycle can be determined by those skilled in the art according to the actual situation, and the above description is only an example and is not limited thereto.

For example, the actual amount demand based on the next period of the latest operation period is obtained as the corresponding historical amount demand interval, which may be, but is not limited to, actual amount demand information of the next period of the latest operation period is determined by querying through a relevant database, a system, a business application, or the like, where the actual amount demand information includes, but is not limited to, an accumulated withdrawal amount of all users of the next period or a historical cash demand estimated in the past, and then one of parameters that can represent the relevant amount demand is selected from the actual amount demand information (for example, the accumulated withdrawal amount or the historical cash demand estimated in the past may be selected), and which interval of a plurality of preset optional intervals the value of the parameter belongs to is determined, and the corresponding interval is determined as the historical amount demand interval. The multiple preset optional intervals may be determined by those skilled in the art according to practical situations, and the embodiment of the present invention is not limited to this, however, any intersection does not occur between the multiple optional intervals, for example, the multiple preset optional intervals include, but are not limited to: [0, 100 ten), [100 ten, 1000 ten), [1000 ten, 5000 ten), [5000 ten, 1 hundred million), and [1 hundred million, + ∞ ]. It should be noted that, for the actual amount demand of the next period based on the latest operation period, a specific implementation manner of obtaining the corresponding historical amount demand interval may be determined by a person skilled in the art according to an actual situation, and the above description is only an example, and does not limit this.

The actual scene of predicting the cash demand is generally to predict the cash demand interval of the next period of the current time, and the process of determining the historical amount demand interval is determined according to the actual amount demand information in the history, so that the accuracy of the determined historical amount demand interval can be improved, the accuracy of the training sample is improved, and the accuracy of the whole cash allocation is improved. In addition, another model predicts the cash demand interval, rather than directly predicting the specific cash demand, because if the granularity is refined to a specific cash demand, the training of the model takes a long time, the execution logic of the model becomes very complex, the time for predicting can be greatly increased, therefore, another model predicts the cash demand interval, the time for overall cash allocation can be greatly reduced, the speed for overall cash allocation is increased, the cash allocation requirement can be met, and the follow-up staff can allocate by themselves only knowing the cash demand interval fed back.

In an alternative embodiment, as shown in fig. 4, the obtaining a predicted cash demand interval of a next cycle of the target banking outlet according to the current business feature vectors of the target banking outlet in multiple cycles and the maturity model includes the following steps:

s401: and obtaining the current input vector according to the current operation characteristic vector and the time of the corresponding period.

S402: and inputting the current input vector into the maturity model so that the maturity model is operated to obtain the predicted cash demand interval.

Illustratively, one current business feature vector corresponds to one cycle. The current operation feature vectors of the multiple cycles may be, but are not limited to, the current operation feature vectors of several cycles closest to the current time, and the number of the current operation feature vectors of the multiple cycles (i.e., the number of the selected cycles) needs to be consistent with the number of the historical operation feature vectors included in one first dimension vector of the historical data matrix. For example, if the current time, 2022, 4, month, 16, is in a period ranging from 2022, 4, month, 12, to 2022, 4, month, 18, the current business feature vectors for multiple periods may include, but are not limited to: an operational feature vector corresponding to a period ranging from 2022 year 4 month 5 day to 2022 year 4 month 11 day, an operational feature vector corresponding to a period ranging from 2022 year 3 month 29 day to 2022 year 4 month 4 day, an operational feature vector corresponding to a period ranging from 2022 year 3 month 22 day to 2022 year 3 month 28 day, and an operational feature vector corresponding to a period ranging from 2022 year 3 month 15 day to 2022 year 3 month 21 day. It should be noted that the current operation feature vector for a plurality of cycles can be determined by those skilled in the art according to practical situations, and the above description is only an example and is not limiting.

For example, the obtaining of the current input vector according to the current operation characteristic vector and the time of the corresponding period may be, but is not limited to, splicing the current operation characteristic vector according to the order of the time of the corresponding period (which may be from early to late or from late to early), and obtaining the current input vector. It should be noted that, for a specific implementation manner of obtaining the current input vector according to the current operation feature vector and the time of the corresponding period, the specific implementation manner may be determined by a person skilled in the art according to an actual situation, and the foregoing description is only an example, and does not limit this.

For example, input data is input into the model to make the model perform operation to obtain a corresponding result, which is a conventional technical means in the art and is not described herein again.

Through the step S401, a plurality of current operation characteristic vectors can be integrated into one current input vector, the number of vectors input into the model is reduced under the condition of not causing input loss, the model is convenient to operate, and the prediction speed of the model is improved. Through the step S402, the model can be used with a correct model using method to obtain the predicted cash demand interval, which is beneficial to determining the speed and accuracy of the predicted cash demand interval, thereby improving the speed and accuracy of the whole cash allocation.

In an alternative embodiment, as shown in fig. 5, the feeding back corresponding warning information to the staff based on the cash reserve amount and the predicted cash demand interval to enable the staff to allocate cash of the target banking outlet includes the following steps:

s501: and judging whether the cash reserve quantity is smaller than the lower limit of the predicted cash demand quantity interval.

S502: if yes, feeding back insufficient cash reserve warning information to the staff according to the cash reserve amount and the predicted cash demand interval, so that the staff allocate cash based on the insufficient cash reserve warning information.

For example, the predicted cash demand interval is one of a plurality of preset optional intervals, which may be determined by those skilled in the art according to practical situations, and this is not limited in this embodiment of the present invention, however, there is no intersection between the plurality of optional intervals, for example, the plurality of preset optional intervals include, but are not limited to: [0, 100 ten), [100 ten, 1000 ten), [1000 ten, 5000 ten), [5000 ten, 1 hundred million), and [1 hundred million, + ∞ ].

Illustratively, the property of the cash reserve amount is the total amount of cash reserved. For example, if the total amount of the reserved cash is 300 ten thousand yuan, the reserved cash amount is 300 ten thousand yuan.

For example, the step of feeding back the warning information of insufficient cash reserve to the staff according to the cash reserve amount and the predicted cash demand interval may be, but is not limited to, feeding back warning information to the staff, such as "the cash demand predicted for a week in the future is in a ten-thousand-yuan to b-ten-thousand-yuan, the current cash reserve amount is c-ten-thousand-yuan, which is smaller than the lower limit of the predicted cash demand range, the cash reserve is insufficient, and the related staff is requested to perform cash allocation to increase the cash reserve", where a refers to the lower limit of the predicted cash demand interval, b refers to the upper limit of the predicted cash demand interval, and c refers to the cash reserve amount. It should be noted that, for a specific implementation manner of feeding back the cash shortage warning information to the staff according to the cash reserve amount and the predicted cash demand interval, the specific implementation manner may be determined by those skilled in the art according to actual situations, and the above description is only an example, and does not limit this.

For example, when the cash reserve is insufficient, the staff member may call the cash from the outside to the inside of the target bank outlet, so as to increase the cash reserve amount to a certain value within the predicted cash demand interval, thereby completing the cash call. It should be noted that the specific blending operation of the working personnel can be determined by the skilled in the art according to the actual situation, and the above description is only an example and is not limiting.

Through the steps, the detailed degree and accuracy of the fed-back warning information can be improved, so that when the cash reserve is insufficient, workers can fully know the related information through the warning information, the cash can be allocated more accurately and rapidly by the workers, the cash withdrawal requirement of a user can be better met, and the operation efficiency of a bank and the experience of the user can be improved.

In an optional embodiment, further comprising:

For example, the step of feeding back the cash reserve excess warning information to the staff according to the cash reserve amount and the predicted cash demand interval may be, but is not limited to, feeding back warning information to the staff, such as "the cash demand amount predicted in the future is in a ten-thousand-yuan to b-ten-thousand-yuan, the current cash reserve amount is c-ten-thousand-yuan, which is greater than the upper limit of the predicted cash demand amount range, the cash reserve is excessive, and the relevant staff is requested to perform cash allocation to reduce the cash reserve, and allocate the deducted cash to other cash reserve shortage websites", where a denotes a lower limit of the predicted cash demand interval, b denotes an upper limit of the predicted cash demand interval, and c denotes the cash reserve amount. It should be noted that, for a specific implementation manner of feeding back the cash reserve excess warning information to the staff according to the cash reserve amount and the predicted cash demand interval, the specific implementation manner may be determined by those skilled in the art according to actual situations, and the above description is only an example, and does not limit this.

For example, when the cash reserve is excessive, the staff may allocate the surplus cash to other banking outlets with insufficient cash reserve, and when the cash reserve amount of the target banking outlet is decreased to a certain value within the predicted cash demand interval, the other banking outlets with insufficient cash reserve may be supplemented with cash in time to complete the cash allocation. It should be noted that the specific blending operation of the working personnel can be determined by the skilled in the art according to the actual situation, and the above description is only an example and is not limiting.

Through the steps, the detailed degree and the accuracy of the fed-back warning information can be improved, so that when the cash reserve volume is excessive, workers can fully know the related information through the warning information, the cash can be allocated more accurately and rapidly by the workers, the cash withdrawal requirement of a user can be better met, and the operation efficiency of a bank and the experience of the user can be improved.

Based on the same principle, the embodiment of the present invention discloses a cash deploying device 600 at a bank outlet, as shown in fig. 6, the cash deploying device 600 at the bank outlet includes:

the training sample generation module 601 is configured to obtain a training sample according to the historical operation feature vectors of the target bank outlets in multiple cycles;

a prediction module 602, configured to train a preset sequence processing model with the training sample to obtain a mature model; obtaining a predicted cash demand interval of the next period of the target bank outlet according to the current operation characteristic vectors of the target bank outlet in a plurality of periods and the mature model;

and the allocating module 603 is configured to determine whether the cash reserve amount of the target bank outlet is in the predicted cash demand interval, and if not, feedback corresponding warning information to the staff based on the cash reserve amount and the predicted cash demand interval, so that the staff allocates cash of the target bank outlet.

In an optional embodiment, the apparatus further comprises a first vector quantization module configured to:

In an optional embodiment, the apparatus further comprises a second quantization module configured to:

In an optional embodiment, the training sample generating module 601 is configured to:

and obtaining the corresponding historical money demand interval based on the actual money demand of the next period of the latest operation period.

In an optional embodiment, the prediction module 602 is configured to:

In an optional embodiment, the allocating module 603 is configured to:

In an optional embodiment, the system further comprises a sub-scheduling module for:

when the cash reserve amount is judged to be not less than the lower limit of the predicted cash demand interval, judging whether the cash reserve amount is greater than the upper limit of the predicted cash demand interval or not;

Since the principle of the cash allocating apparatus 600 for banking outlets for solving the problems is similar to the above method, the implementation of the cash allocating apparatus 600 for banking outlets can refer to the implementation of the above method, and will not be described herein again.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. A typical implementation device is a computer device, which may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

In a typical example, the computer device comprises in particular a memory, a processor and a computer program stored on the memory and executable on the processor, which when executed by the processor implements the method as described above.

Referring now to FIG. 7, shown is a schematic block diagram of a computer device 700 suitable for use in implementing embodiments of the present application.

As shown in fig. 7, the computer device 700 includes a Central Processing Unit (CPU) 701, which can perform various appropriate works and processes according to a program stored in a Read Only Memory (ROM) 702 or a program loaded from a storage section 708 into a Random Access Memory (RAM) 703. In the RAM703, various programs and data necessary for the operation of the system 700 are also stored. The CPU701, ROM702, and RAM703 are connected to each other via a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.

The following components are connected to the I/O interface 705: an input portion 706 including a keyboard, a mouse, and the like; an output section 707 including components such as a Cathode Ray Tube (CRT), a liquid crystal feedback (LCD), and the like, and a speaker; a storage section 708 including a hard disk and the like; and a communication section 709 including a network interface card such as a LAN card, a modem, or the like. The communication section 709 performs communication processing via a network such as the internet. A drive 710 is also connected to the I/O interface 705 as needed. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 710 as necessary, so that a computer program read out therefrom is mounted as necessary in the storage section 708.

In particular, according to an embodiment of the present invention, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the invention include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 709, and/or installed from the removable medium 711.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functionality of the units may be implemented in one or more software and/or hardware when implementing the present application.

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.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of other like elements in a process, method, article, or apparatus comprising the element.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application 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 so forth) having computer-usable program code embodied therein.

The application may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The application may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, as for the system embodiment, since it is substantially similar to the method embodiment, the description is relatively simple, and reference may be made to the partial description of the method embodiment for relevant points.

The above description is only an example of the present application and is not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A cash allocating method for a bank outlet is characterized by comprising the following steps:

and judging whether the cash reserve volume of the target bank outlet is in the predicted cash demand volume interval, if not, feeding corresponding warning information back to the staff based on the cash reserve volume and the predicted cash demand volume interval so that the staff allocates cash of the target bank outlet.

2. The method of claim 1, further comprising:

3. The method of claim 1, further comprising:

4. The method as claimed in claim 1, wherein the obtaining of the training sample according to the historical operation feature vector of the target banking outlet for a plurality of periods comprises:

5. The method of claim 4, wherein the stitching the current historical operation matrix and the plurality of misalignment matrices along a first dimension direction to obtain a historical data matrix comprises:

6. The method of claim 4, wherein obtaining training samples based on the first dimension vector of the historical data matrix comprises:

7. The method of claim 6, wherein the obtaining the corresponding historical monetary demand intervals based on the plurality of first dimension vectors of the historical data matrix comprises:

8. The method as claimed in claim 6, wherein the obtaining the predicted cash demand interval of the next cycle of the target banking outlet according to the current business feature vector of the target banking outlet in a plurality of cycles and the maturity model comprises:

9. The method of claim 1, wherein the feeding back corresponding warning information to the staff based on the cash reserve amount and the predicted cash demand interval to enable the staff to allocate cash of the target bank outlet comprises:

judging whether the cash reserve volume is smaller than the lower limit of the predicted cash demand volume interval or not;

10. The method of claim 9, further comprising:

11. A cash dispenser for banking outlets, comprising:

12. A computer arrangement comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 1-10 when executing the program.

13. A computer-readable medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-10.

14. A computer program product, characterized in that the computer program product comprises a computer program which, when being executed by a processor, carries out the method of any one of claims 1 to 10.