CN112950004A - Enterprise financial risk early warning method and device and computer readable storage medium - Google Patents
Enterprise financial risk early warning method and device and computer readable storage medium Download PDFInfo
- Publication number
- CN112950004A CN112950004A CN202110169278.1A CN202110169278A CN112950004A CN 112950004 A CN112950004 A CN 112950004A CN 202110169278 A CN202110169278 A CN 202110169278A CN 112950004 A CN112950004 A CN 112950004A
- Authority
- CN
- China
- Prior art keywords
- early warning
- financial
- data
- enterprise
- model
- 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
- 238000000034 method Methods 0.000 title claims abstract description 54
- 230000006403 short-term memory Effects 0.000 claims abstract description 27
- 238000003062 neural network model Methods 0.000 claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 16
- 230000007787 long-term memory Effects 0.000 claims abstract description 15
- 238000000605 extraction Methods 0.000 claims abstract description 6
- 230000004927 fusion Effects 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 47
- 230000015654 memory Effects 0.000 claims description 22
- 230000000694 effects Effects 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 description 22
- 238000013528 artificial neural network Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000000502 dialysis Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q40/12—Accounting
- G06Q40/125—Finance or payroll
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Educational Administration (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Software Systems (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Tourism & Hospitality (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Technology Law (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses an enterprise financial risk early warning method, an enterprise financial risk early warning device and a storage medium, wherein index data in a sample enterprise historical financial statement is selected; normalizing the index data; newly adding a historical data feature fusion channel to the long and short term memory neural network model to obtain an ultra-long and short term memory neural network model; performing feature extraction on the normalized index data by adopting a long-term and short-term memory neural network model, and after outputting features, arranging a full connection layer for judging whether the enterprise has financial risks or not in a secondary classification manner; and selecting sample data, and giving sample weight for training to obtain an early warning model. According to the invention, from the three reports, the specific contents of a plurality of financial indexes which can be subjected to risk prediction and are determined by experts, the long-term and short-term memory neural network model of the core prediction model, and the financial early warning accuracy is improved by training the long-term and short-term memory neural network model and adding the prediction model formed by full-connection layers.
Description
Technical Field
The invention belongs to the field of financial risk early warning, and particularly relates to an enterprise financial risk early warning method and device and a computer readable storage medium.
Background
Financial risk is the variability of the equity return and the risk that a business may lose its ability to repay. As debt, lease and priority equity funding increase in the capital structure of an enterprise, the capital costs of an enterprise increase, and as a result, the likelihood of an enterprise losing cash back increases. Another aspect of corporate financial risk relates to the relative dispersion of the gains made available to stockholders. In summary, the financial risk of a business encompasses the variability of the stakeholder's future benefits and the potential for the business to lose its ability to repay. Both of these aspects are directly related to the business risk of the enterprise, i.e., the expected business revenue variation.
Currently, the financial risk early warning methods generally include: and the quantitative trend method predicts the financial crisis of the company by utilizing the trend of indexes such as net profits, stockholder equity and equity/liability. The disadvantage is that complex connection between financial indexes of the dialysis company cannot be carried out, and the prediction effect is poor. The discrimination model method adopts Logistic regression model, Fisher discrimination model, Bayes discrimination model, Z-score model, second-class linear discrimination model and the like to predict the financial crisis of the company. The method has the disadvantages that the whole financial condition of a company is difficult to accurately depict, and the accuracy is not high.
Due to the complexity of financial risk sources, the traditional model and method cannot accurately describe the inherent law of the traditional model and method, so that the risk early warning accuracy rate is low.
Disclosure of Invention
In order to solve the problem that the traditional model and method cannot accurately describe the inherent law of the traditional model and method due to the complexity of financial risk sources, so that the risk early warning accuracy rate is low, the invention aims to provide an enterprise financial risk early warning method, device and computer-readable storage medium, so that the accuracy rate of enterprise financial risk early warning is improved, and an important reference basis is provided for enterprise development.
In a first aspect, the invention provides an enterprise financial risk early warning method, which includes the following steps:
selecting index data in a sample enterprise historical financial statement; the index data at least comprises three types of index items, wherein the three types of index items are a profit statement characteristic index, an asset and debt statement characteristic index and a cash flow statement characteristic index;
normalizing the index data;
newly adding a historical data feature fusion channel to the long and short term memory neural network model to obtain an ultra-long and short term memory neural network model;
performing feature extraction on the normalized index data by adopting the ultra-long short-term memory neural network model;
a full connection layer is arranged behind the output characteristics and used for judging whether the enterprise has financial risks in a secondary classification mode to obtain a primary early warning model;
selecting sample data, giving a weight corresponding to the sample data, and training a primary preliminary early warning model by adopting the sample data and the weight to obtain an early warning model;
and inputting index data of the financial information of the enterprise to be evaluated into the early warning model to obtain an evaluation result.
According to the technology, the specific contents of multiple financial indexes for risk prediction are calculated from the three reports, the core prediction model long-short term memory neural network model is used, and the financial early warning accuracy is improved through training of the long-short term memory neural network model and increasing of the prediction model formed by full-connection layers.
In one possible design, the profit sheet characteristic indicators include at least business income, business cost, business profit, total profit, income tax rate, net profit, and basic revenue per share indicators;
the balance sheet characteristic indicators include monetary funds, accounts receivable, inventory, liquidity aggregate, fixed net balance, total of assets, liquidity aggregate, and non-liquidity aggregate indicators;
the cash flow table characteristic indexes comprise debt total, stockholder equity total, beginning cash and cash equivalent balance, cash flow net amount generated by operation activity, cash flow net amount generated by investment activity, cash flow net amount generated by financing activity, cash and cash equivalent net increase amount and end cash and cash equivalent balance indexes.
In one possible design, the index data is normalized, and all index data is converted into data between 0 and 1 by the following formula:
x is the index data, and x is the index data,
the normalized index data is obtained.
In one possible design, the long-short term memory neural network model is augmented with the following model:
F1＝Relu(Xt)*T(t-1)；
F2＝F1+Relu(σ*Xt)；
ht＝F2*F3*F4；
λ is a coefficient from 0 to 1; relu is a standard Relu activation function, tanh is a tanh activation function, and sigma is a Sigmoid activation function; xtFor t years of data input, htFor the training result output of T years, TtFor intermediate output, F3 ═ X (σ ═ X)t)*tanh(F5)，F4＝λ*tanh(Xt)，F5＝tanh(Xt)*(σ*Xt)+(σ*Xt)*Relu(Xt)*Tt-1。
In one possible design, the ultralong short-term memory neural network model has multiple inputs and outputs in common.
In one possible design, the method for the ultralong short-term memory neural network model to share multiple inputs and outputs is as follows:
when index data in a sample enterprise historical financial statement are acquired, a sample data matrix is formed according to a time period;
presetting a weight coefficient matrix corresponding to a corresponding time period;
inputting a sub-matrix formed by 1 st row to nth row of all columns of the sample data matrix for the first time, wherein n is an integer greater than 1, and the sub-matrix represents financial information in a period of time; inner products of each line of the submatrix and the weight coefficient matrix a1 of the corresponding time period, and the obtained data is the standard output of the first-time model;
the second input is the result of the inner product of the standard output of the first model, each row of a submatrix formed by the (n + 1) th row to the (2 n) th row of the sample data matrix and the weight coefficient matrix a2 of the corresponding time period, and the obtained data is the standard output of the second model;
the mth input is the standard output of the mth model, the result of the inner product of each row of the submatrix formed by the (m-1) n +1 row to the m x n row of the sample data matrix and the coefficient matrix a (m) of the corresponding time period, m is an integer larger than 2, and the obtained data is the standard output of the mth model;
until m × n is the total row number of the sample data matrix; and after the standard output of the mth model and the weight coefficient matrix a corresponding to all the time are subjected to inner product, inputting a standard full-connection layer to obtain a two-dimensional output result of the risk probability and the non-risk probability, judging that the financial risk exists when the risk probability is greater than the non-risk probability, and judging that the financial risk does not exist when the risk probability is less than the non-risk probability.
In one possible design, the weight fraction corresponding to the sub-matrix that is further away from the present time is smaller.
In one possible design, the sample data is financial data of enterprises without financial risk and enterprises with financial risk, and for enterprises with financial risk, enterprises which are firstly named in the current year are selected.
In a second aspect, the present invention provides an enterprise financial risk early warning device, comprising a memory, a processor and a transceiver connected in sequence, wherein the memory is used for storing a computer program, the transceiver is used for transmitting and receiving messages, and the processor is used for reading the computer program and executing the method of the first aspect or any one of the possible designs of the first aspect.
In a third aspect, the invention provides a computer-readable storage medium having stored thereon instructions which, when run on a computer, perform the method as set forth in the first aspect or any one of the possible designs of the first aspect.
In a fourth aspect, the present invention provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method as described above in the first aspect or any one of the possible designs of the first aspect.
The invention has the beneficial effects that:
the method extracts the specific contents of a plurality of financial indexes for risk prediction from three reports, the sequence and the weight of the financial indexes input into the model, and the design structure of the long-term and short-term memory neural network model of the core prediction model improves the financial early warning accuracy through a prediction model formed by a training mode of a multi-time long-term and short-term memory neural network model, a weight coefficient matrix and a full connection layer.
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 steps of an enterprise financial risk early warning method provided by the present invention.
FIG. 2 is a diagram of an MLSTM model in an embodiment provided by the present invention.
FIG. 3 is a schematic diagram of an overall structure of a prediction model according to an embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and specific embodiments. It should be noted that the description of the embodiments is provided to help understanding of the present invention, but the present invention is not limited thereto. Specific structural and functional details disclosed herein are merely illustrative of example embodiments of the invention. This invention may, however, be embodied in many alternate forms and should not be construed as limited to the embodiments set forth herein.
It will be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments of the present invention.
It should be understood that, for the term "and/or" as may appear herein, it is merely an associative relationship that describes an associated object, meaning that three relationships may exist, e.g., a and/or B may mean: a exists alone, B exists alone, and A and B exist at the same time; for the term "/and" as may appear herein, which describes another associative object relationship, it means that two relationships may exist, e.g., a/and B, may mean: a exists independently, and A and B exist independently; in addition, for the character "/" that may appear herein, it generally means that the former and latter associated objects are in an "or" relationship.
It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments of the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises," "comprising," "includes" and/or "including," when used herein, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, numbers, steps, operations, elements, components, and/or groups thereof.
It should also be noted that, in some alternative designs, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed substantially concurrently, or the figures may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
It should be understood that specific details are provided in the following description to facilitate a thorough understanding of example embodiments. However, it will be understood by those of ordinary skill in the art that the example embodiments may be practiced without these specific details. For example, systems may be shown in block diagrams in order not to obscure the examples in unnecessary detail. In other instances, well-known processes, structures and techniques may be shown without unnecessary detail in order to avoid obscuring example embodiments.
As shown in fig. 1, a first aspect of this embodiment provides an enterprise financial risk early warning method, including the following steps:
selecting index data in a sample enterprise historical financial statement; the index data at least comprises three types of index items, wherein the three types of index items are a profit statement characteristic index, an asset and debt statement characteristic index and a cash flow statement characteristic index; when the indexes are selected, the financial conditions of the enterprises appearing in the market are contained in financial statements of the enterprises, and because the financial statements comprehensively and systematically disclose the operation results and the cash flow of the enterprises in a certain period, the profitability, the debt paying capacity, the investment income, the development prospect and the like of the enterprises are shown. Meanwhile, the financial risk is also hidden in the financial statement of the enterprise, and risk assessment about the enterprise can be obtained by analyzing some indexes in the financial statement.
Normalizing the index data; in order to eliminate the inconsistency of the measuring units of the financial indexes, which may cause the unequal length reduction of the model gradient, the data are normalized; converting all index data into data between 0 and 1 by adopting the following formula:
x is the index data, and x is the index data,
the normalized index data is obtained.
Newly adding a historical data feature fusion channel to the long and short term memory neural network model to obtain an ultra-long and short term memory neural network model; because the financial risk prediction related to the invention is based on historical timeline data, the invention provides a very long short-term memory neural network (MLSTM) based on a long short-term memory neural network (LSTM), which takes the hidden layer output at the previous moment and the data at the current moment as the input of the moment, and the unique network structure enables the data at the historical moment to have influence on the output at the current moment;
performing feature extraction on the normalized index data by adopting the ultralong short-term memory neural network model, and after outputting features, arranging a full connection layer for secondary classification judgment on whether the enterprise has financial risks or not to obtain a primary early warning model; in specific implementation, the second classification is used for judging whether the enterprise has financial risk, namely judging whether the enterprise has financial risk or not.
Selecting sample data, and inputting the sample data into the preliminary early warning model for training to obtain an early warning model; since the stock market in China has an ST system, namely the stock of an enterprise covered by ST or ST indicates that the enterprise encounters a financial crisis, the invention represents the financial risk of the enterprise through ST. The risk early warning is generally carried out on enterprises without risks at present, namely enterprises without the risk of ST, and has no significance of risk early warning on enterprises with the risk of ST, so that the method is used for training a model which inputs selected sample data items meeting conditions into step three.
And inputting index data of the financial information of the enterprise to be evaluated into the early warning model to obtain an evaluation result.
In one possible design, the final 23 criteria in step one: the related three reports of profit statement, balance statement and cash flow statement. For a certain enterprise needing prediction, each index of the enterprise selects data of every quarter in the past 5 years, namely 20 data in total. Then 23 metrics for 460 items of data. These 460 data form a matrix with time of day, 20 rows, the first row being the data in the first quarter 5 years ago, the columns of the matrix being the indices, 23 columns, the order of the indices being in the order in Table 1. The characteristic indexes of the profit list at least comprise business income, business cost, business profit, total profit, income tax expense, net profit and basic income indexes of each share;
the balance sheet characteristic indicators include monetary funds, accounts receivable, inventory, liquidity aggregate, fixed net balance, total of assets, liquidity aggregate, and non-liquidity aggregate indicators;
the cash flow table characteristic indexes comprise debt total, stockholder equity total, beginning cash and cash equivalent balance, cash flow net amount generated by operation activity, cash flow net amount generated by investment activity, cash flow net amount generated by financing activity, cash and cash equivalent net increase amount and end cash and cash equivalent balance indexes.
Performing feature extraction on data corresponding to 23 indexes in the financial statement by adopting an ultra-long short-term memory network model, wherein the ultra-long short-term memory network model is shown in figure 2 and represented by black and thick lines as a newly added model; firstly, adding a historical data feature further extraction channel; the channel is a channel F1 which,
F1＝Relu(Xt)*T(t-1)；
the mathematical expression for the intermediate data value F2 is:
F2＝F1+Relu(σ*Xt)；
secondly, the method comprises the following steps: newly adding a historical data feature fusion channel (the lower half part in the figure is added with a thick part data flow): because past operational benefits of enterprises directly influence the current financial condition, namely historical data in the financial statement directly influence data in the current financial statement, a newly-added historical data fusion channel is provided, and the mathematical expression of the channel is as follows:
F4＝λ*tanh(Xt)；
and (3) fusing an intermediate output result F3 of the channel and the traditional long and short term memory neural network and newly added further extracted channel intermediate data F2 to obtain the output of the improved long and short term memory neural network (MLSTM), wherein the mathematical expression is as follows: h ist＝F2*F3*F4；
The mathematical expression of the intermediate output result F3 of the conventional long-short term memory neural network is as follows:
F3＝(σ*Xt)*tanh(F5)，
F5＝tanh(Xt)*(σ*Xt)+(σ*Xt)*Relu(Xt)*Tt-1。
λ is a coefficient from 0 to 1; relu is a standard Relu activation function, tanh is a tanh activation function, and sigma is a Sigmoid activation function; xtFor t years of data input, htFor the training result output of T years, TtIs the intermediate output.
Based on the MLSTM core model, all process models of the method in the third step are shown in FIG. 2, and the ultra-long short-term memory neural network model has multiple inputs and outputs.
In one possible design, the method for the ultralong short-term memory neural network model to share multiple inputs and outputs is as follows:
when index data in a sample enterprise historical financial statement are acquired, a sample data matrix is formed according to a time period;
presetting a weight coefficient matrix corresponding to a corresponding time period;
inputting a sub-matrix formed by 1 st row to nth row of all columns of the sample data matrix for the first time, wherein n is an integer greater than 1, and the sub-matrix represents financial information in a period of time; inner products of each line of the submatrix and the weight coefficient matrix a1 of the corresponding time period, and the obtained data is the standard output of the first-time model;
the second input is the result of the inner product of the standard output of the first model, each row of a submatrix formed by the (n + 1) th row to the (2 n) th row of the sample data matrix and the weight coefficient matrix a2 of the corresponding time period, and the obtained data is the standard output of the second model;
the mth input is the standard output of the mth model, the result of the inner product of each row of the submatrix formed by the (m-1) n +1 row to the m x n row of the sample data matrix and the coefficient matrix a (m) of the corresponding time period, m is an integer larger than 2, and the obtained data is the standard output of the mth model;
until m × n is the total row number of the sample data matrix; and after the standard output of the mth model and the weight coefficient matrix a corresponding to all the time are subjected to inner product, inputting a standard full-connection layer to obtain a two-dimensional output result of the risk probability and the non-risk probability, judging that the financial risk exists when the risk probability is greater than the non-risk probability, and judging that the financial risk does not exist when the risk probability is less than the non-risk probability.
Specific examples are given below, and as shown in fig. 3, this step has 5 inputs and outputs:
the first input is a sub-matrix formed by the 1 st row to the 4 th row of all the columns of the matrix in the step one, each row of the sub-matrix is inner-multiplied by a coefficient matrix a1, and the obtained data is the standard output of the MLSTM. The second input is the standard output of the first MLSTM, the result of the inner product of each row of the submatrix formed by the 5 th row to the 8 th row of the matrix in the step one and the coefficient a2, and the obtained data is the standard output of the MLSTM. And so on until 20 rows of data are input into the model.
And finally, the data obtained in the 5 th time is the standard output of the MLSTM and is an output vector hn with 23 columns, and after the hn vector is subjected to inner product with the coefficient a, the data is input into a standard full-connection layer to obtain a two-dimensional output result: and the risk probability and the non-risk probability are real numbers in the middle of 0 to 1, and when the risk probability is greater than the non-risk probability, the financial risk is judged to be present, and when the risk probability is less than the non-risk probability, the financial risk is judged not to be present.
Wherein the weight coefficient matrixes a, a1, a2, a3, a4 and a5 are respectively:
a1＝[0.1，0.2,0.3，0.1,0.2，0.3,0.3，0.2，0.1，0.3，0.2，0.1，0.3，0.2，0.1，0.1，0.2，0.1，0.3，0.1，0.3，0.1，0.3]
a2＝[0.4，0.2,0.4，0.2,0.4，0.2,0.3，0.2，0.4，0.3，0.4，0.1，0.2，0.3，0.1，0.4，0.3，0.1，0.3，0.2，0.3，0.4，0.3]
a3＝[0.5，0.1,0.5，0.5,0.5，0.2,0.4，0.5，0.4，0.5，0.3，0.5，0.4，0.5，0.3，0.5，0.3，0.5，0.3，0.5，0.4，0.4，0.4]
a4＝[0.6，0.5,0.3，0.6,0.5，0.5,0.5，0.6，0.3，0.6，0.3，0.4，0.3，0.4，0.5，0.4，0.5，0.1，0.5，0.5，0.3，0.5，0.3]
a5＝[0.7，0.8,0.6，0.6,0.8，0.6,0.6，0.6，0.6，0.7，0.7，0.8，0.8，0.7，0.3，0.7，0.8，0.7，0.6，0.7，0.8，0.6，0.7]
a＝[0.9，0.8,0.8，0.8,0.9，0.9,0.8，0.9，0.8，0.8，0.8，0.9，0.9，0.8，0.9，0.8，0.9，0.9，0.9，0.9，0.9，0.9，0.8]。
in one possible design, the weight fraction corresponding to the sub-matrix that is further away from the present time is smaller. Since the longer the time is, the lower the referential of the data is, the longer the data is, the lower the weight is set, and the accuracy is improved.
In one possible design, the sample data is financial data of enterprises without financial risk and enterprises with financial risk, and for enterprises with financial risk, enterprises which are firstly named in the current year are selected. In step four, for rigor of training, for enterprises without financial risk, enterprises which have never been named ST in the last 5 years are selected, and for financial risk enterprises, enterprises which are named ST for the first time in this year are selected. After the above conditions are screened, a total of 500 risk-free enterprises and 260 financial risk-existing enterprises in the training examples of the invention select and train a model in which a total of 419520 sample data items are input into step three from 23 items in each quarter from 2014 to 2020.
According to the invention, from three reports, the specific contents of a plurality of financial indexes which can be subjected to risk prediction and are determined by experts, the sequence and the weight of the financial indexes input into the model, and the design structure of the MLSTM model of the core prediction model improve the financial early warning accuracy through a prediction model formed by a multi-time MLSTM network training mode, a weight coefficient matrix and a full connection layer.
Selecting a model training data mode: for financial risk enterprises, enterprises that have never been ST in the last 5 years, enterprises that are first named ST this year, are selected. For non-financial risk enterprises, enterprises that have never been ST in the past 5 years, and enterprises that have not been named ST this year are selected.
In the present example, the effectiveness of the proposed method is verified by testing the financial status of 200 risk-free listed enterprises and 100 financial risk-present listed enterprises, and the prediction results are shown in the following table.
Enterprise financial crisis prediction result
| Predicted financial crisis occurrence in 2020 | Predicting non-occurrence of financial crisis in 2020 | |
| Actual financial crisis in 2020 | 85 | 15 |
| Financial crisis does not actually occur in 2020 | 14 | 186 |
In the table, 85 enterprises which are predicted to have financial crises in 2020 actually have financial crises, and 14 enterprises which are predicted to have financial crises in 2020 do not have financial crises; 200 risk-free listed enterprises are China, and 15 enterprises in the enterprises predicted not to have financial crises in 2020 actually have financial crises; 186 actually no financial crisis occurred;
it can be seen that the accuracy of the method for predicting whether an enterprise is involved in a financial crisis reaches 90.33%, and the recall rate reaches 85.85%. Therefore, the method can accurately predict the financial risk of the enterprise, and has high practicability.
In a second aspect, the present invention provides an enterprise financial risk early warning device, comprising a memory, a processor and a transceiver connected in sequence, wherein the memory is used for storing a computer program, the transceiver is used for transmitting and receiving messages, and the processor is used for reading the computer program and executing the method of the first aspect or any one of the possible designs of the first aspect.
For the working process, working details and technical effects of the foregoing apparatus provided in the second aspect of this embodiment, reference may be made to the interaction method described in any one of the above first aspect or the first aspect, which is not described herein again.
A third aspect of the present embodiment provides another enterprise financial risk early warning apparatus for implementing the enterprise financial risk early warning method in any one of the first aspect or the first aspect, including a memory, a processor, and a transceiver, which are communicatively connected in sequence, wherein the memory is used for storing a computer program, the transceiver is used for transceiving a message, and the processor is used for reading the computer program and executing the steps as implemented in any one of the first aspect or the first aspect. For example, the Memory may include, but is not limited to, a Random-Access Memory (RAM), a Read-Only Memory (ROM), a Flash Memory (Flash Memory), a First-in First-out (FIFO), and/or a First-in Last-out (FILO), and the like; the processor may not be limited to the use of a microprocessor model number STM32F105 family; the transceiver may be, but is not limited to, a Wireless Fidelity (WiFi) Wireless transceiver, a bluetooth Wireless transceiver, a General Packet Radio Service (GPRS) Wireless transceiver, a ZigBee Wireless transceiver (ieee 802.15.4 standard-based low power local area network protocol), and/or a ZigBee Wireless transceiver. In addition, the enterprise financial risk early warning device can also include but not be limited to including power module, display screen and other necessary parts.
The working process, working details and technical effects of the enterprise financial risk early warning device provided in the third aspect of this embodiment may refer to the interaction method described in the first aspect or any one of the possible designs in the first aspect, which is not described herein again.
A fourth aspect of the present invention provides a computer-readable storage medium storing instructions for implementing any one of the possible designs of the enterprise financial risk early warning method according to the first aspect or the first aspect, where the instructions are stored on the computer-readable storage medium, and when the instructions are executed on a computer, the method is implemented as any one of the possible designs of the enterprise financial risk early warning device according to the first aspect or the first aspect. The computer-readable storage medium refers to a carrier for storing data, and may include, but is not limited to, floppy disks, optical disks, hard disks, flash memories, flash disks and/or Memory sticks (Memory sticks), etc., and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
For a working process, working details, and technical effects of the foregoing computer-readable storage medium provided in the fourth aspect of this embodiment, reference may be made to the interaction method in any one of the above first aspect or the first aspect, which is not described herein again.
A fifth aspect of the present embodiments provides a computer program product comprising instructions which, when executed on a computer, cause the computer to perform a method for enterprise financial risk early warning as described in the first aspect or any one of the possible designs of the first aspect. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable devices.
The embodiments described above are merely illustrative, and may or may not be physically separate, if referring to units illustrated as separate components; if reference is made to a component displayed as a unit, it may or may not be a physical unit, and may be located in one place or distributed over a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: modifications may be made to the embodiments described above, or equivalents may be substituted for some of the features described. And such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Finally, it should be noted that the present invention is not limited to the above alternative embodiments, and that various other forms of products can be obtained by anyone in light of the present invention. The above detailed description should not be taken as limiting the scope of the invention, which is defined in the claims, and which the description is intended to be interpreted accordingly.
Claims (10)
1. An enterprise financial risk early warning method is characterized by comprising the following steps:
selecting index data in a sample enterprise historical financial statement; the index data at least comprises three types of index items, wherein the three types of index items are a profit statement characteristic index, an asset and debt statement characteristic index and a cash flow statement characteristic index;
normalizing the index data;
newly adding a historical data feature fusion channel to the long and short term memory neural network model to obtain an ultra-long and short term memory neural network model;
performing feature extraction on the normalized index data by adopting the ultra-long short-term memory neural network model;
a full connection layer is arranged behind the output characteristics and used for judging whether the enterprise has financial risks in a secondary classification mode to obtain a primary early warning model;
selecting sample data, giving a weight corresponding to the sample data, and training a preliminary early warning model by adopting the sample data and the weight to obtain an early warning model;
and inputting index data of the financial information of the enterprise to be evaluated into the early warning model to obtain an evaluation result.
2. The enterprise financial risk early warning method according to claim 1,
the characteristic indexes of the profit list at least comprise business income, business cost, business profit, total profit, income tax expense, net profit and basic income indexes of each share;
the balance sheet characteristic indicators include monetary funds, accounts receivable, inventory, liquidity aggregate, fixed net balance, total of assets, liquidity aggregate, and non-liquidity aggregate indicators;
the cash flow table characteristic indexes comprise debt total, stockholder equity total, beginning cash and cash equivalent balance, cash flow net amount generated by operation activity, cash flow net amount generated by investment activity, cash flow net amount generated by financing activity, cash and cash equivalent net increase amount and end cash and cash equivalent balance indexes.
3. An enterprise financial risk early warning method according to claim 1, wherein the index data is normalized, and all index data is converted into data between 0 and 1 by adopting the following formula:
x is the index data, and x is the index data,
the normalized index data is obtained.
4. The enterprise financial risk early warning method according to claim 1, wherein the long-short term memory neural network model is additionally provided with the following models:
F1＝Relu(Xt)*T(t-1)；
F2＝F1+Relu(σ*Xt)；
ht＝F2*F3*F4；
λ is a coefficient from 0 to 1; relu is a standard Relu activation function, tanh is a tanh activation function, and sigma is a Sigmoid activation function; xtFor t years of data input, htFor the training result output of T years, TtIntermediate output for training results in the t year, wherein F3 ═ Xt)*tanh(F5)，F4＝λ*tanh(Xt)，F5＝tanh(Xt)*(σ*Xt)+(σ*Xt)*Relu(Xt)*Tt-1。
5. The enterprise financial risk early warning method according to claim 4, wherein the ultralong short-term memory neural network model has multiple inputs and outputs.
6. The enterprise financial risk early warning method according to claim 5, wherein the method for the long-short term memory neural network model to have multiple inputs and outputs is as follows:
when index data in a sample enterprise historical financial statement are acquired, a sample data matrix is formed according to a time period;
presetting a weight coefficient matrix corresponding to a corresponding time period;
inputting a sub-matrix formed by 1 st row to nth row of all columns of the sample data matrix for the first time, wherein n is an integer greater than 1, and the sub-matrix represents financial information in a period of time; inner products of each line of the submatrix and the weight coefficient matrix a1 of the corresponding time period, and the obtained data is the standard output of the first-time model;
the second input is the result of the inner product of the standard output of the first model, each row of a submatrix formed by the (n + 1) th row to the (2 n) th row of the sample data matrix and the weight coefficient matrix a2 of the corresponding time period, and the obtained data is the standard output of the second model;
the mth input is the standard output of the mth model, the result of the inner product of each row of the submatrix formed by the (m-1) n +1 row to the m x n row of the sample data matrix and the coefficient matrix a (m) of the corresponding time period, m is an integer larger than 2, and the obtained data is the standard output of the mth model;
until m × n is the total row number of the sample data matrix; and after the standard output of the mth model and the weight coefficient matrix a corresponding to all the time are subjected to inner product, inputting a standard full-connection layer to obtain a two-dimensional output result of the risk probability and the non-risk probability, judging that the financial risk exists when the risk probability is greater than the non-risk probability, and judging that the financial risk does not exist when the risk probability is less than the non-risk probability.
7. The enterprise financial risk early warning method according to claim 6, wherein the weight proportion corresponding to the submatrix which is longer from the current time is smaller.
8. The enterprise financial risk early warning method according to claim 1, wherein the sample data is financial data of enterprises without financial risk and enterprises with financial risk, and for enterprises with financial risk, enterprises which are firstly named and risked within the current year are selected.
9. The utility model provides an enterprise financial risk early warning device which characterized in that: the enterprise financial risk early warning method comprises a memory, a processor and a transceiver which are connected in sequence, wherein the memory is used for storing a computer program, the transceiver is used for transmitting and receiving messages, and the processor is used for reading the computer program and executing the enterprise financial risk early warning method according to any one of claims 1-8.
10. A computer-readable storage medium characterized by: the computer-readable storage medium stores instructions which, when executed on a computer, perform the enterprise financial risk early warning method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110169278.1A CN112950004A (en) | 2021-02-07 | 2021-02-07 | Enterprise financial risk early warning method and device and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110169278.1A CN112950004A (en) | 2021-02-07 | 2021-02-07 | Enterprise financial risk early warning method and device and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112950004A true CN112950004A (en) | 2021-06-11 |
Family
ID=76243265
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110169278.1A Pending CN112950004A (en) | 2021-02-07 | 2021-02-07 | Enterprise financial risk early warning method and device and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112950004A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114118526A (en) * | 2021-10-29 | 2022-03-01 | 中国建设银行股份有限公司 | Enterprise risk prediction method, device, equipment and storage medium |
| CN114154866A (en) * | 2021-12-02 | 2022-03-08 | 北京顶象技术有限公司 | Marketing enterprise financial risk early warning method and system |
| CN117009907A (en) * | 2023-08-31 | 2023-11-07 | 国网江苏省电力有限公司宿迁供电分公司 | Financial system early warning method and system based on trend abnormal value detection |
| CN117422573A (en) * | 2023-10-30 | 2024-01-19 | 广州金财智链数字科技有限公司 | Financial tax data prediction method and system based on dynamic neural network |
| CN117541411A (en) * | 2023-11-08 | 2024-02-09 | 厦门工学院 | Construction method and system of financial prediction model based on big data processing |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108876166A (en) * | 2018-06-27 | 2018-11-23 | 平安科技（深圳）有限公司 | Financial risk authentication processing method, device, computer equipment and storage medium |
| CN110490720A (en) * | 2019-07-08 | 2019-11-22 | 深圳壹账通智能科技有限公司 | Financial data analysis and early warning method, apparatus, computer equipment and storage medium |
| CN111292007A (en) * | 2020-02-28 | 2020-06-16 | 中国工商银行股份有限公司 | Supplier financial risk prediction method and device |
| CN111915218A (en) * | 2020-08-14 | 2020-11-10 | 中国工商银行股份有限公司 | Financial counterfeiting identification method and system based on LSTM-CNN |
-
2021
- 2021-02-07 CN CN202110169278.1A patent/CN112950004A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108876166A (en) * | 2018-06-27 | 2018-11-23 | 平安科技（深圳）有限公司 | Financial risk authentication processing method, device, computer equipment and storage medium |
| CN110490720A (en) * | 2019-07-08 | 2019-11-22 | 深圳壹账通智能科技有限公司 | Financial data analysis and early warning method, apparatus, computer equipment and storage medium |
| CN111292007A (en) * | 2020-02-28 | 2020-06-16 | 中国工商银行股份有限公司 | Supplier financial risk prediction method and device |
| CN111915218A (en) * | 2020-08-14 | 2020-11-10 | 中国工商银行股份有限公司 | Financial counterfeiting identification method and system based on LSTM-CNN |
Non-Patent Citations (1)
| Title |
|---|
| 肖毅等: "基于TEI@I方法论的企业财务风险预警模型研究", 《管理评论》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114118526A (en) * | 2021-10-29 | 2022-03-01 | 中国建设银行股份有限公司 | Enterprise risk prediction method, device, equipment and storage medium |
| CN114154866A (en) * | 2021-12-02 | 2022-03-08 | 北京顶象技术有限公司 | Marketing enterprise financial risk early warning method and system |
| CN117009907A (en) * | 2023-08-31 | 2023-11-07 | 国网江苏省电力有限公司宿迁供电分公司 | Financial system early warning method and system based on trend abnormal value detection |
| CN117009907B (en) * | 2023-08-31 | 2024-07-19 | 国网江苏省电力有限公司宿迁供电分公司 | Financial system early warning method and system based on trend abnormal value detection |
| CN117422573A (en) * | 2023-10-30 | 2024-01-19 | 广州金财智链数字科技有限公司 | Financial tax data prediction method and system based on dynamic neural network |
| CN117541411A (en) * | 2023-11-08 | 2024-02-09 | 厦门工学院 | Construction method and system of financial prediction model based on big data processing |
| CN117541411B (en) * | 2023-11-08 | 2024-06-11 | 厦门工学院 | Construction method and system of financial prediction model based on big data processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112950004A (en) | Enterprise financial risk early warning method and device and computer readable storage medium | |
| Li et al. | Machine learning and credit ratings prediction in the age of fourth industrial revolution | |
| Gupta et al. | Impact of market-based finance on SMEs failure | |
| Caggiano et al. | Early warning systems and systemic banking crises in low income countries: A multinomial logit approach | |
| US10453142B2 (en) | System and method for modeling and quantifying regulatory capital, key risk indicators, probability of default, exposure at default, loss given default, liquidity ratios, and value at risk, within the areas of asset liability management, credit risk, market risk, operational risk, and liquidity risk for banks | |
| Demyanyk et al. | Financial crises and bank failures: A review of prediction methods | |
| Chiu et al. | The analysis of Taiwanese bank efficiency: Incorporating both external environment risk and internal risk | |
| Magerman et al. | Heterogeneous firms and the micro origins of aggregate fluctuations | |
| Cocco et al. | Aging in place, housing maintenance, and reverse mortgages | |
| Mizen et al. | Forecasting US bond default ratings allowing for previous and initial state dependence in an ordered probit model | |
| WO2015094545A1 (en) | System and method for modeling and quantifying regulatory capital, key risk indicators, probability of default, exposure at default, loss given default, liquidity ratios, and value at risk, within the areas of asset liability management, credit risk, market risk, operational risk, and liquidity risk for banks | |
| Markellos et al. | Sovereign debt markets in light of the shadow economy | |
| Do et al. | Predicting loss severities for residential mortgage loans: A three-step selection approach | |
| WO2021257610A1 (en) | Time series forecasting and visualization methods and systems | |
| Sanders et al. | Noise trader sentiment in futures markets | |
| Tsioumas et al. | Capturing the impact of economic forces on the dry bulk freight market | |
| Lagasio et al. | Assessing bank default determinants via machine learning | |
| Kim et al. | Predicting corporate defaults using machine learning with geometric-lag variables | |
| Calabrese et al. | Generalized extreme value regression for binary rare events data: an application to credit defaults | |
| Hunter et al. | Aggregate economy risk and company failure: An examination of UK quoted firms in the early 1990s | |
| Pratap et al. | Macroeconomic effects of uncertainty: a Google trends-based analysis for India | |
| Zhao et al. | Unintended consequences of tax incentives on firms' human capital composition: Evidence from China | |
| López-Iturriaga et al. | Predicting bankruptcy using neural networks in the current financial crisis: A study of US commercial banks | |
| Tang | Application of Nonlinear Autoregressive with Exogenous Input (NARX) neural network in macroeconomic forecasting, national goal setting and global competitiveness assessment | |
| Chancharat et al. | Firms in financial distress, a survival model analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210611 |