Method for intelligently realizing big data early warning based on model strategy

Abstract

The invention discloses a method for intelligently realizing big data early warning based on a model strategy, and relates to the field of financial data analysis, and the method comprises the steps: S1, collecting historical financial data; s2, preprocessing the historical financial data; s3, numerical value filling and numerical value replacement; s4, risk early warning; by identifying and judging a missing numerical value position and a numerical value abnormal position in the data and selecting the most suitable filling value and replacement value for filling and replacement according to specific conditions, the reasonability and accuracy of filling the data are improved; according to the method, the current financial data and the historical correction report data are compared and analyzed, the risk value of each income and expenditure type is calculated, potential financial risks and income or expenditure abnormal changes can be identified in real time, necessary early warning and adjustment measures are taken in time, and the intelligence and agility of enterprise financial management are improved.

Description

Technical Field

[0001] The present invention relates to the field of financial data early warning, and in particular to a method for realizing big data early warning based on model strategy intelligence. Background Art

[0002] Turnover and operating costs are very important in corporate financial data. They have a key impact and role on the company's operations and financial status. Turnover directly reflects the scale and effectiveness of the company's sales activities. Efficient management and increase in turnover can promote the company's revenue growth and market share expansion. Operating costs show the cost level of the company's production or service provision. Effective cost control helps to improve gross profit margins and profitability. Therefore, it is necessary to conduct big data early warnings on turnover and operating costs in financial data;

[0003] At present, early warning of financial data still relies heavily on manual analysis and intervention, with a low degree of automation, which not only increases the workload but also leads to human errors. Summary of the invention

[0004] The purpose of the present invention is to solve the problem of financial data and to propose a method for realizing big data early warning based on model strategy intelligence.

[0005] The purpose of the present invention can be achieved by the following technical solution: A method for realizing big data early warning based on model strategy intelligence, comprising the following steps:

[0006] S1: Collect historical financial data and classify the historical financial data into various types;

[0007] S2: Preprocess the historical financial data and find out the missing and abnormal value locations;

[0008] S3: Fill in missing and abnormal positions and replace values;

[0009] S4: By comparing and analyzing the current financial data with the correction report, identify whether there is a risk and generate corresponding warnings; the specific steps are:

[0010] Acquire current financial data, wherein the current financial data includes: current operating income information and current operating cost information; the current operating income information includes the income type and the current income amount corresponding to each income type, and the current operating cost information includes the expenditure type and the current expenditure amount corresponding to each expenditure type;

[0011] The historical income amount in the correction income report is recorded as Tri, where r = 1, 2, 3 ... R, R is a positive integer, R represents the total type of the correction income report (total income type), and r represents the serial number of any one of the correction income report types; the historical expenditure amount in the correction expense report is recorded as Qhj, where h = 1, 2, 3 ... H, H is a positive integer, H represents the total type of the correction expense report (total expenditure type), and h represents the serial number of any one of the correction expense report types;

[0012] Substitute the current income amount value Cr, the current expenditure amount value Kr, the historical income amount value Tri and the historical expenditure amount value Qhj into the model strategy formula group 1 The income risk value YCr corresponding to each income type and the expenditure risk value YKr corresponding to each expenditure type are calculated; wherein b1 and b2 are set correction factors, and b1 and b2 are custom values; the income risk value YCr corresponding to each income type is compared with the set income warning value, and when the income risk value YCr is greater than the set income warning value, an income warning signal is generated, wherein the income warning signal includes the income type and the income amount corresponding to the income type; the expenditure risk value YKr corresponding to each expenditure type is compared with the set expenditure warning value, and when the expenditure risk value YKr is greater than the set expenditure warning value, an expenditure warning signal is generated, wherein the expenditure warning signal includes the expenditure type and the expenditure amount corresponding to the expenditure type;

[0013] As a preferred embodiment of the present invention, the specific steps of preprocessing the historical financial data are as follows:

[0014] Identify duplicate information in historical operating income information and historical operating cost information respectively, and remove duplicate information to ensure the uniqueness of each piece of information;

[0015] Select any one of the income types, and generate a historical income report of the income type by chronological order for each historical income amount, thereby obtaining a historical income report of each income type; select any one of the expenditure types, and generate a historical expenditure report of the expenditure type by chronological order for each historical expenditure amount, thereby obtaining a historical income report of each expenditure type;

[0016] Identify locations in each historical income statement and each historical expense statement where data is missing and record them as missing value locations;

[0017] Take any one of the historical income reports, obtain the historical income amount value Fi corresponding to each time, and calculate the average of the historical income amount corresponding to each time to get the historical income average Where i = 1, 2, 3, ..., I, where I is a positive integer, I represents the total number of time points of historical income amounts, and i represents the sequence number of any time point of the historical income amounts;

[0018] Take any one of the historical expenditure reports, obtain the historical expenditure amount value Lj corresponding to each time, and calculate the average of the historical expenditure amount corresponding to each time to obtain the historical expenditure mean Where J = 1, 2, 3 ... J, J is a positive integer, J represents the total number of time points of historical expenditure amounts, and j represents the sequence number of any time point of the historical expenditure amounts;

[0019] The historical income amount Fi and the historical income mean Historical expenditure amount value Lj and historical expenditure mean Substitute into model strategy formula group 2 The income deviation value Pi and the expenditure deviation value Pj at each time point are calculated respectively; a1 and a2 are the set weight factors, and a1 and a2 are custom values;

[0020] An income deviation threshold is set. When the income deviation value Pi is greater than the income deviation threshold, the income deviation value is abnormal and is recorded as a preliminary abnormal income deviation value. The preliminary abnormal income deviation value is identified, and the historical income amount at that time point is recorded as an abnormal income amount, and the position is recorded as a numerical abnormal position. An expenditure deviation threshold is set. When the expenditure deviation value Pj is greater than the expenditure deviation threshold, the expenditure deviation value is abnormal and is recorded as a preliminary abnormal expenditure deviation value. The preliminary abnormal expenditure deviation value is identified, and the historical income amount at that time point is recorded as an abnormal income amount, and the position is recorded as a numerical abnormal position.

[0021] As a preferred embodiment of the present invention, the specific steps of filling the missing value positions are:

[0022] When the value type of the missing position is income type, the operating historical income amount corresponding to the time point corresponding to the missing position is extracted as the center, and several time points adjacent to the center (time points before and after the center) are taken as the relevant income amount of the missing position, thereby obtaining the relevant income amount corresponding to any missing position;

[0023] Take any missing value position, calculate the mean of the relevant income amount of the missing value position to obtain the filling value one; extract the relevant income amount of the missing value position, sort it from large to small, and obtain the median, which is the filling value two; extract the relevant income amount of the missing value position and perform mode statistics to obtain the filling value three; if the operating income amount of each corresponding time point has no mode, calculate the mean of the operating income amount of each corresponding time point, which is the filling value three;

[0024] When two or more of the filling values ​​1, 2 and 3 are equal, the equal filling values ​​are recorded as the target filling values; when the filling values ​​1, 2 and 3 are not equal, the filling values ​​1, 2 and 3 are averaged to obtain the target filling value; and the target filling value is filled to the corresponding missing value position;

[0025] Similarly, when the value type of the missing value position is an expenditure type, the target filling value of each expenditure type can be obtained;

[0026] As a preferred embodiment of the present invention, the specific steps of replacing the value of the abnormal position are:

[0027] When the value type of the abnormal value position is income type, the operating income amount corresponding to the time point corresponding to the abnormal value position is extracted as the center, and several time points adjacent to the center (time points before and after the center) are taken as the relevant income amount of the abnormal value position, thereby obtaining the relevant income amount corresponding to any abnormal value position;

[0028] Take any abnormal value position, calculate the mean of the relevant income amount at the abnormal value position to obtain replacement value seven; extract the relevant income amount at the abnormal value position, sort it from large to small, and obtain the median, which is the replacement value eight; extract the relevant income amount at the abnormal value position and perform mode statistics to obtain replacement value nine; if the operating income amount at each corresponding time point has no mode, calculate the mean of the operating income amount at each corresponding time point, which is the replacement value nine;

[0029] When two or more replacement values ​​among replacement value seven, replacement value eight and replacement value nine are equal, the equal replacement values ​​are recorded as target replacement values; when replacement value seven, replacement value eight and replacement value nine are not equal, the replacement value seven, replacement value eight and replacement value nine are averaged to obtain the target replacement value; and the target replacement value replaces the value at the corresponding abnormal position;

[0030] Similarly, when the value type of the abnormal value position is an expenditure type, the target replacement value of each expenditure type can be obtained;

[0031] The historical income report corresponding to each income type and the historical expenditure report corresponding to each expenditure type are replaced by filling in the missing and abnormal positions to obtain the corrected income report corresponding to each income type and the corrected expenditure report corresponding to each expenditure type.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. By identifying and judging the missing value positions and abnormal value positions in the data, using a variety of methods such as mean, sorted median and mode to calculate the filling value, and selecting the most suitable filling value and replacement value according to the specific situation, and then filling and replacing the values ​​of the missing value positions and abnormal value positions, the rationality and accuracy of the filled data are improved, which provides a solid foundation for subsequent data analysis, prediction and early warning identification.

[0034] 2. By comparing and analyzing current financial data and historical correction report data and calculating the risk value of each type of income and expenditure, it is possible to identify potential financial risks and abnormal changes in income or expenditure in real time, so as to take necessary early warning and adjustment measures in time, thereby improving the intelligence and agility of corporate financial management.

[0035] In summary, automated data comparison and risk identification reduce the need for manual intervention, making the financial management process more efficient and accurate; at the same time, real-time monitoring and early warning systems enable companies to respond quickly to changes, improving the flexibility and adaptability of financial management. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0037] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0038] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] See also Figure 1 As shown, a method for realizing big data early warning based on model strategy intelligence includes the following steps:

[0040] S1: Collect historical financial data and store them in a database; the historical financial data includes historical operating income information and historical operating cost information; the historical operating income information includes income types and historical income amounts corresponding to each income type, and the historical operating cost information includes expenditure types and historical expenditure amounts corresponding to each expenditure type;

[0041] S2: Preprocess the historical financial data. The specific steps of preprocessing are as follows:

[0042] Identify duplicate information in historical operating income information and historical operating cost information respectively, and remove duplicate information to ensure the uniqueness of each piece of information;

[0043] Select any one of the income types, and generate a historical income report of the income type by chronological order for each historical income amount, thereby obtaining a historical income report of each income type; select any one of the expenditure types, and generate a historical expenditure report of the expenditure type by chronological order for each historical expenditure amount, thereby obtaining a historical income report of each expenditure type;

[0044] Identify locations in each historical income statement and each historical expense statement where data is missing and record them as missing value locations;

[0045] Take any one of the historical income statements, record the historical income amount corresponding to each time as Fi, and calculate the average of the historical income amount corresponding to each time to get the historical income average as Where i = 1, 2, 3, ..., I, where I is a positive integer, I represents the total number of time points of historical income amounts, and i represents the sequence number of any time point of the historical income amounts;

[0046] Take any one of the historical expenditure reports, record the historical expenditure amount corresponding to each time as Lj, and calculate the average of the historical expenditure amount corresponding to each time to obtain the historical expenditure average as Where J = 1, 2, 3 ... J, J is a positive integer, J represents the total number of time points of historical expenditure amounts, and j represents the sequence number of any time point of the historical expenditure amounts;

[0047] The historical income amount Fi and the historical income mean Historical expenditure amount Lj and historical expenditure mean Substitute the set model strategy formula group 1 The income deviation value Pi and the expenditure deviation value Pj at each time point are calculated respectively; a1 and a2 are the set weight factors, a1 is 2.34, and a2 is 2.56;

[0048] Set an income deviation threshold. When the income deviation value Pi is greater than the income deviation threshold, the income deviation value Pi is abnormal and recorded as a preliminary abnormal income deviation value. Identify the preliminary abnormal income deviation value, record the historical income amount at that time point as the abnormal income amount, and record the position as the numerical abnormal position. Set an expenditure deviation threshold. When the expenditure deviation value Pi is greater than the expenditure deviation threshold, the expenditure deviation value Pi is abnormal and recorded as a preliminary abnormal expenditure deviation value. Identify the preliminary abnormal expenditure deviation value, record the historical income amount at that time point as the abnormal income amount, and record the position as the numerical abnormal position.

[0049] S3: Extract the time points and value types corresponding to the missing value positions and the abnormal value positions. It should be noted that the value types corresponding to the missing value positions and the abnormal value positions refer to the income type and the historical income amount corresponding to each income type, the expenditure type and the historical expenditure amount corresponding to each expenditure type; fill the missing value positions and the abnormal value positions and replace the values, specifically:

[0050] When the value type of the missing position is income type, the operating historical income amount corresponding to the time point corresponding to the missing position is extracted as the center, and several time points adjacent to the center (time points before and after the center) are taken as the relevant income amount of the missing position, thereby obtaining the relevant income amount corresponding to any missing position;

[0051] Take any position where the value is missing, calculate the mean of the relevant income amount at the position where the value is missing to obtain the filling value one; extract the relevant income amount at the position where the value is missing, sort it from large to small, and obtain the median, which is the filling value two; extract the relevant income amount at the position where the value is missing to perform mode statistics, and obtain the filling value three; it should be noted that when performing mode statistics on the operating income amount, the value needs to be rounded off to facilitate statistics. If the operating income amount at each corresponding time point has no mode, the operating income amount at each corresponding time point is calculated as the mean, which is the filling value three;

[0052] When two or more of the filling values ​​1, 2 and 3 are equal, the equal filling values ​​are recorded as the target filling values; when the filling values ​​1, 2 and 3 are not equal, the filling values ​​1, 2 and 3 are averaged to obtain the target filling value; and the target filling value is filled to the corresponding missing value position;

[0053] When the value type of the missing position is expenditure type, the operating expenditure amount corresponding to the time point corresponding to the missing position is extracted as the center, and several time points adjacent to the center (time points before and after the center) are taken as the relevant expenditure amount of the missing position, thereby obtaining the relevant expenditure amount corresponding to any missing position;

[0054] Take any position where the value is missing, calculate the mean of the relevant expenditure amount at the position where the value is missing to obtain the filling value four; extract the relevant expenditure amount at the position where the value is missing. Sort from large to small to obtain the median, which is the filling value five; extract the relevant expenditure amount at the position where the value is missing to perform mode statistics, and obtain the filling value six; it should be noted that when performing mode statistics on the operating expenditure amount, the value needs to be rounded to the nearest hundred to facilitate statistics. If the operating expenditure amount at each corresponding time point has no mode, the operating expenditure amount at each corresponding time point is calculated by the mean, which is the filling value six;

[0055] When two or more of the filling values ​​4, 5 and 6 are equal, the equal filling values ​​are recorded as the target filling values; when the filling values ​​4, 5 and 6 are not equal, the filling values ​​4, 5 and 6 are averaged to obtain the target filling value; and the target filling value is filled to the corresponding missing value position;

[0056] When the value type of the abnormal value position is income type, the operating income amount corresponding to the time point corresponding to the abnormal value position is extracted as the center, and several time points adjacent to the center (time points before and after the center) are taken as the relevant income amount of the abnormal value position, thereby obtaining the relevant income amount corresponding to any abnormal value position;

[0057] Take any abnormal value position, calculate the mean of the relevant income amount at the abnormal value position to obtain replacement value seven; extract the relevant income amount at the abnormal value position, sort it from large to small, and obtain the median, which is the replacement value eight; extract the relevant income amount at the abnormal value position and perform mode statistics to obtain replacement value nine; it should be noted that when performing mode statistics on the operating income amount, the value needs to be rounded off to facilitate statistics. If the operating income amount at each corresponding time point has no mode, the operating income amount at each corresponding time point is calculated by the mean, which is the replacement value nine;

[0058] When two or more of the replacement values ​​7, 8 and 9 are equal in filling value, the equal filling values ​​are recorded as the target filling values; when the replacement values ​​7, 8 and 9 are not equal, the replacement values ​​7, 8 and 9 are averaged to obtain the target filling value; and the target filling value replaces the value at the corresponding abnormal position;

[0059] When the value type of the abnormal value position is the expenditure type, the operating expenditure amount corresponding to the time point corresponding to the abnormal value position is extracted as the center, and several time points adjacent to the center (time points before and after the center) are taken as the relevant expenditure amount of the abnormal value position, thereby obtaining the relevant expenditure amount corresponding to any abnormal value position;

[0060] Take any abnormal value position, calculate the mean of the relevant expenditure amount at the abnormal value position to obtain the replacement value of ten; extract the relevant expenditure amount at the abnormal value position, sort them from large to small, and obtain the median, which is the replacement value eleven; extract the relevant expenditure amount at the abnormal value position and perform mode statistics to obtain the replacement value twelve; it should be noted that when performing mode statistics on the operating expenditure amount, the value needs to be rounded to the nearest hundred to facilitate statistics. If the operating expenditure amount at each corresponding time point has no mode, the operating expenditure amount at each corresponding time point is calculated by the mean, which is the replacement value twelve;

[0061] When two or more of the filling values ​​10, 11 and 12 are equal, the equal filling values ​​are recorded as the target filling values; when the replacement value 10, 11 and 12 are not equal, the replacement value 10, 11 and 12 are averaged to obtain the target replacement value; and the target replacement value replaces the value at the corresponding abnormal position;

[0062] The historical income report corresponding to each income type and the historical expenditure report corresponding to each expenditure type are replaced by filling in the missing positions and abnormal positions to obtain the corrected income report corresponding to each income type and the corrected expenditure report corresponding to each expenditure type;

[0063] By identifying and judging the missing value positions and abnormal value positions in the data, using a variety of methods such as mean, sorted median and mode to calculate the filling value, and selecting the most suitable filling value and replacement value according to the specific situation, and then filling and replacing the values ​​in the missing value positions and abnormal value positions, the rationality and accuracy of the filled data are improved, providing a solid foundation for subsequent data analysis, prediction and early warning identification.

[0064] S4: By comparing and analyzing the current financial data with the correction income report and correction expenditure report, identify whether there are risks and generate corresponding warnings; specifically:

[0065] Obtain current financial data, where the current financial data includes: current operating income information and current operating cost information; the current operating income information includes the income type and the current income amount corresponding to each income type, recorded as Cr, and the current operating cost information includes the expenditure type and the current expenditure amount corresponding to each expenditure type, recorded as Kr;

[0066] The historical income amount in the correction income report is recorded as Tri, where r = 1, 2, 3 ... R, R is a positive integer, R represents the total type of the correction income report (total income type), and r represents the serial number of any one of the correction income report types; the historical expenditure amount in the correction expense report is recorded as Qhj, where h = 1, 2, 3 ... H, H is a positive integer, H represents the total type of the correction expense report (total expenditure type), and h represents the serial number of any one of the correction expense report types;

[0067] Substitute the current income amount Cr, the current expenditure amount Kr, the historical income amount as Tri and the historical expenditure amount as Qhj into the set model strategy formula group 2 Obtain the income risk value YCr corresponding to each income type and the expenditure risk value YKr corresponding to each expenditure type; wherein b1 and b2 are set correction factors, b1 takes a value of 1.12, and b2 takes a value of 1.08; compare the income risk value YCr corresponding to each income type with the set income warning value, and when the income risk value YCr is greater than the set income warning value, generate an income warning signal, wherein the income warning signal includes the income type and the income amount corresponding to the income type; compare the expenditure risk value YKr corresponding to each expenditure type with the set expenditure warning value, and when the expenditure risk value YKr is greater than the set expenditure warning value, generate an expenditure warning signal, wherein the expenditure warning signal includes the expenditure type and the expenditure amount corresponding to the expenditure type;

[0068] By comparing and analyzing current financial data and historical correction report data and calculating the risk value of each type of income and expenditure, it is possible to identify potential financial risks and abnormal changes in income or expenditure in real time, so that necessary early warning and adjustment measures can be taken in time, improving the intelligence and agility of corporate financial management.

[0069] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for realizing big data early warning based on model strategy intelligence, characterized in that: The following steps are involved: S1: Collect historical financial data and classify the historical financial data into various types; S2: Preprocess the historical financial data and find out the missing and abnormal value locations; S3: Fill in missing and abnormal positions and replace values; S4: By comparing and analyzing the current financial data with the correction report, identify whether there is risk and generate corresponding warnings; The specific steps are: Acquire current financial data, wherein the current financial data includes: current operating income information and current operating cost information; the current operating income information includes the income type and the current income amount corresponding to each income type, and the current operating cost information includes the expenditure type and the current expenditure amount corresponding to each expenditure type; The current income amount, the current expenditure amount, the historical income amount and the historical expenditure amount are normalized to obtain the income risk value corresponding to each income type and the expenditure risk value corresponding to each expenditure type; the income risk value corresponding to each income type is compared with the set income warning value. When the income risk value is greater than the set income warning value, an income warning signal is generated, wherein the income warning signal includes the income type and the income amount corresponding to the income type; the expenditure risk value corresponding to each expenditure type is compared with the set expenditure warning value. When the income risk value is greater than the set expenditure warning value, an expenditure warning signal is generated, wherein the expenditure warning signal includes the expenditure type and the expenditure amount corresponding to the expenditure type.

2. According to claim 1, a method for intelligently implementing big data early warning based on model strategy is characterized in that: The specific steps for preprocessing historical financial data are: Identify duplicate information in historical operating income information and historical operating cost information respectively, and remove duplicate information to ensure the uniqueness of each piece of information; Select any one of the income types, and generate a historical income report of the income type by chronological order for each historical income amount, thereby obtaining a historical income report of each income type; select any one of the expenditure types, and generate a historical expenditure report of the expenditure type by chronological order for each historical expenditure amount, thereby obtaining a historical income report of each expenditure type; Identify locations in each historical income statement and each historical expense statement where data is missing and record them as missing value locations; Take any one of the historical income statements and historical expenditure statements to obtain the corresponding historical income amount value, historical income mean, historical expenditure amount value and historical expenditure mean, and normalize them to obtain the income deviation value and expenditure deviation value at each time point respectively; Set an income deviation threshold. When the income deviation value is greater than the income deviation threshold, the income deviation value is abnormal and recorded as the preliminary abnormal income deviation value. Identify the initial abnormal income deviation value, record the historical income amount at that time point as the abnormal income amount, and record the position as the abnormal value position; set an expenditure deviation threshold, when the expenditure deviation value is greater than the expenditure deviation threshold, the expenditure deviation value is abnormal and recorded as the initial abnormal expenditure deviation value; Identify the preliminary abnormal expenditure deviation value, record the historical income amount at that point in time as the abnormal income amount, and record the position as the numerical abnormal position.

3. According to claim 1, a method for intelligently implementing big data early warning based on model strategy is characterized in that: The specific steps to fill in the missing values ​​are: When the value type of the missing position is income type, the operating historical income amount corresponding to several time points adjacent to the time point corresponding to the missing position is extracted as the relevant income amount of the missing position, thereby obtaining the relevant income amount corresponding to any missing position; Take any missing value position, calculate the mean of the relevant income amount of the missing value position to obtain the filling value one; extract the relevant income amount of the missing value position, sort it from large to small, and obtain the median, which is the filling value two; extract the relevant income amount of the missing value position and perform mode statistics to obtain the filling value three; if the operating income amount of each corresponding time point has no mode, calculate the mean of the operating income amount of each corresponding time point, which is the filling value three; When two or more of the filling values ​​1, 2 and 3 are equal, the equal filling values ​​are recorded as the target filling values; when the filling values ​​1, 2 and 3 are not equal, the filling values ​​1, 2 and 3 are averaged to obtain the target filling value; and the target filling value is filled to the corresponding missing value position; Similarly, when the value type of the missing value position is an expenditure type, the target filling value of each expenditure type can be obtained.

4. The method for realizing big data early warning based on model strategy intelligence according to claim 3 is characterized in that: The specific steps for replacing the numerical values ​​of abnormal positions are: When the value type of the abnormal value position is income type, the operating income amount corresponding to several time points adjacent to the time point corresponding to the abnormal value position is extracted as the relevant income amount of the abnormal value position, thereby obtaining the relevant income amount corresponding to any abnormal value position; Take any abnormal value position, calculate the mean of the relevant income amount at the abnormal value position to obtain replacement value seven; extract the relevant income amount at the abnormal value position, sort it from large to small, and obtain the median, which is the replacement value eight; extract the relevant income amount at the abnormal value position and perform mode statistics to obtain replacement value nine; if the operating income amount at each corresponding time point has no mode, calculate the mean of the operating income amount at each corresponding time point, which is the replacement value nine; When two or more replacement values ​​among replacement value seven, replacement value eight and replacement value nine are equal, the equal replacement values ​​are recorded as target replacement values; when replacement value seven, replacement value eight and replacement value nine are not equal, the replacement value seven, replacement value eight and replacement value nine are averaged to obtain the target replacement value; and Replace the value at the position of the corresponding abnormal value with the target replacement value; Similarly, when the value type of the abnormal value position is an expenditure type, the target replacement value of each expenditure type can be obtained; The historical income report corresponding to each income type and the historical expenditure report corresponding to each expenditure type are replaced by filling in the missing and abnormal positions to obtain the corrected income report corresponding to each income type and the corrected expenditure report corresponding to each expenditure type.

Images