CN113723710B - Customer loss prediction method, system, storage medium and electronic equipment - Google Patents
Customer loss prediction method, system, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN113723710B CN113723710B CN202111068135.8A CN202111068135A CN113723710B CN 113723710 B CN113723710 B CN 113723710B CN 202111068135 A CN202111068135 A CN 202111068135A CN 113723710 B CN113723710 B CN 113723710B
- Authority
- CN
- China
- Prior art keywords
- client
- customer
- predicted
- function value
- target
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 230000006870 function Effects 0.000 claims description 220
- 230000006399 behavior Effects 0.000 claims description 89
- 238000011156 evaluation Methods 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 14
- 230000002159 abnormal effect Effects 0.000 claims description 10
- 230000000694 effects Effects 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 7
- 238000004590 computer program Methods 0.000 description 6
- 230000009471 action Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q30/00—Commerce
- G06Q30/01—Customer relationship services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Technology Law (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention provides a customer loss prediction method, a system, a storage medium and electronic equipment, which can be applied to the field of artificial intelligence or the field of finance. The method comprises the following steps: determining a first function value and a second function value of a client to be predicted; selecting a client to be predicted as a first target client, and judging whether a second target client exists in the client set to be predicted, wherein a first function value of the second target client is larger than a first function value of the first target client, and a second function value of the second target client is larger than a second function value of the first target client; if the second target client exists, judging whether a third target client exists, wherein the first function value of the third target client is smaller than the first function value of the first target client, and the second function value of the first target client is smaller than the second function value of the first target client; if the third target client does not exist, the first target client is determined to be the first potential attrition client. The invention can improve the accuracy of customer loss prediction.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method and system for predicting customer loss, a storage medium, and an electronic device.
Background
As market competition increases, the cost of maintaining old customers is greatly reduced compared to the cost of developing new customers. Therefore, maintaining a relationship with an old customer avoids customer churn, which is of great importance to the enterprise. At present, the enterprise often analyzes the characteristics of the lost clients according to the experience of actual service personnel to obtain a prediction result, and the determined potential lost clients are inaccurate due to insufficient service knowledge of the service personnel and subjective judgment of the service personnel.
Disclosure of Invention
The embodiment of the invention aims to provide a customer loss prediction method, a system, a storage medium and electronic equipment, so as to improve the accuracy of customer loss prediction. The specific technical scheme is as follows:
the invention provides a customer churn prediction method, which comprises the following steps:
acquiring a client behavior data of each client to be predicted in a client set to be predicted; the customer behavior data comprises first behavior data and second behavior data; the first behavior data reflects the activity degree of the client, and the second behavior data reflects the satisfaction degree of the client;
determining a first objective function according to the characteristic type of the first behavior data, and determining a second objective function according to the characteristic type of the second behavior data;
determining a first function value and a second function value of each client to be predicted; the first function value is obtained by inputting the first row of data into the first objective function; the second function value is obtained by inputting the second behavior data into the second objective function;
selecting one to-be-predicted client from the to-be-predicted client set as a first target client, and judging whether a second target client exists in the to-be-predicted client set, wherein a first function value of the second target client is larger than a first function value of the first target client, and a second function value of the second target client is larger than a second function value of the first target client;
if the second target client exists, judging whether a third target client exists in the client set to be predicted, wherein a first function value of the third target client is smaller than a first function value of the first target client, and a second function value of the first target client is smaller than a second function value of the first target client;
and if the third target client does not exist, determining that the first target client is a first potential attrition client.
Optionally, after selecting one to-be-predicted client from the to-be-predicted client set as the first target client, the method further includes:
determining the maximum value and the minimum value of the first function values of the clients to be predicted in the client set to be predicted, and obtaining the maximum first function value and the minimum first function value; determining the maximum value and the minimum value of the second function values of the clients to be predicted in the client set to be predicted, and obtaining the maximum second function value and the minimum second function value;
and determining that the first target client is a first potential loss client when the first function value of the first target client is the maximum first function value and the second function value of the first target client is the minimum second function value or when the first function value of the first target client is the minimum first function value and the second function value of the first target client is the maximum second function value.
Optionally, the method further comprises:
if the second target client does not exist or the third target client exists, the first target client is stored in a first target client set; the clients in the first set of target clients do not include the first potential attrition client;
selecting a first target client from the first target client set as a fourth target client, and judging whether a fifth target client exists in the client set to be predicted, wherein a first function value of the fifth target client is smaller than a first function value of the fourth target client, and a second function value of the fifth target client is smaller than a second function value of the fourth target client;
if the fifth target client exists, determining the number of the fifth target clients, and if the number of the fifth target clients is 1, determining the fourth target client as a second potential loss client; the customer churn probability of the second potential churn customer is less than the customer churn probability of the first potential churn customer.
Optionally, after the determining that the fourth target client is a second potential attrition client, the method further comprises:
and deleting the first target client and the fourth target client in the client to be predicted set, and taking the rest clients to be predicted as sixth target clients, wherein the client loss probability of the sixth target clients is smaller than the client loss probability of the second potential loss clients.
Optionally, after obtaining the set of clients to be predicted and the client behavior data of each client to be predicted in the set of clients to be predicted, the method further comprises:
judging whether the feature types of clients to be predicted in the client set to be predicted are the same or not;
and if the feature types of the clients to be predicted are different, carrying out unified processing on the feature types of the clients to be predicted.
Optionally, after obtaining the set of clients to be predicted and the client behavior data of each client to be predicted in the set of clients to be predicted, the method further comprises:
judging whether the client behavior data corresponding to the characteristic type of the client to be predicted in the client set to be predicted is abnormal or not;
if so, replacing the abnormal customer behavior data with preset data.
Optionally, after obtaining the set of clients to be predicted and the client behavior data of each client to be predicted in the set of clients to be predicted, the method further comprises:
and carrying out normalization processing on the client behavior data to obtain normalized client behavior data.
The invention also provides a customer churn prediction system, comprising:
the client data acquisition module is used for acquiring a client set to be predicted and client behavior data of each client to be predicted in the client set to be predicted; the customer behavior data comprises first behavior data and second behavior data; the first behavior data reflects the activity degree of the client, and the second behavior data reflects the satisfaction degree of the client;
the objective function determining module is used for determining a first objective function according to the characteristic type of the first behavior data and determining a second objective function according to the characteristic type of the second behavior data;
the function value determining module is used for determining a first function value and a second function value of each client to be predicted; the first function value is obtained by inputting the first row of data into the first objective function; the second function value is obtained by inputting the second behavior data into the second objective function;
the first judging module is used for selecting one to-be-predicted client from the to-be-predicted client set as a first target client and judging whether a second target client exists in the to-be-predicted client set, wherein a first function value of the second target client is larger than a first function value of the first target client, and a second function value of the second target client is larger than a second function value of the first target client;
the second judging module is used for judging whether a third target client exists in the client set to be predicted or not when the second target client exists, wherein the first function value of the third target client is smaller than the first function value of the first target client, and the second function value of the first target client is smaller than the second function value of the first target client;
and the first customer churn prediction module is used for determining that the first target customer is a first potential churn customer when the third target customer does not exist.
The present invention also provides a computer readable storage medium having a program stored thereon, which when executed by a processor, implements the customer churn prediction method described above.
The present invention also provides an electronic device including:
at least one processor, and at least one memory, bus, connected to the processor;
the processor and the memory complete communication with each other through the bus; the processor is configured to invoke the program instructions in the memory to perform the customer churn prediction method described above.
According to the client churn prediction method, the client churn prediction system, the storage medium and the electronic equipment provided by the embodiment of the invention, a first objective function is determined according to the characteristic type of first behavior data, and a second objective function is determined according to the characteristic type of second behavior data; the first behavior data reflects the activity degree of the client, and the second behavior data reflects the satisfaction degree of the client; inputting the first row of data into a first objective function to obtain a first function value; inputting the second behavior data into a second objective function to obtain a second function value; the method has the advantages that the activity degree of the client can be reflected through the first function value, the satisfaction degree of the client can be reflected through the second function value, and whether the client has loss possibility can be determined through the first function value and the second function value, so that client loss prediction is realized, and the accuracy of client loss prediction can be improved.
Of course, it is not necessary for any one product or method of practicing the invention to achieve all of the advantages set forth above at the same time.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a client churn prediction method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a relationship between predicted client function values according to an embodiment of the present invention;
FIG. 3 is a block diagram of a customer churn prediction system according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention provides a customer churn prediction method, as shown in figure 1, which comprises the following steps:
step 101: and acquiring the client behavior data of each client to be predicted in the client set to be predicted. The customer behavior data includes first behavior data and second behavior data; the first behavior data reflects the liveness of the customer and the second behavior data reflects the satisfaction of the customer.
When customer churn prediction is performed, the customers to be predicted in the customer set to be predicted can be selected as bank customers. Collecting silverWhen the line client acts on data, the first line data reflects the activity degree of the client, and the first line data comprises estimated annual income A= { a of the client 1 ,a 2 ,…,a n Annual average deposit b= { B }, of customers 1 ,b 2 ,…,b n Number of years of customer use c= { C } 1 ,c 2 ,…,c n Average number of transactions d= { D }, client year 1 ,d 2 ,…,d n Customer possession product quantity e= { E } 1 ,e 2 ,…,e n }. The second behavior data reflects the satisfaction of the customer, and the second behavior data includes service evaluation data G= { G on the customer line 1 ,g 2 ,…,g n ' offline service evaluation data k= { K 1 ,k 2 ,…,k n }. n represents the number of clients to be predicted, and the nth client to be predicted has a n 、b n 、c n 、d n 、e n 、g n 、k n A total of 7 eigenvalues.
After step 101, the customer churn prediction method further includes:
judging whether the feature types of clients to be predicted in the client set to be predicted are the same or not; and if the feature types of the clients to be predicted are different, carrying out unified processing on the feature types of the clients to be predicted.
If the client to be predicted lacks one or more of the 7 feature values, the median value of the feature in all clients to be predicted is replaced for the missing feature value.
After step 101, the customer churn prediction method further includes:
judging whether the client behavior data corresponding to the characteristic type of the client to be predicted in the client set to be predicted is abnormal or not; if so, replacing the abnormal customer behavior data with preset data.
If the characteristic value of the client to be predicted is abnormal, the abnormal value is replaced by a fixed value 1.
After step 101, the customer churn prediction method further includes:
and carrying out normalization processing on the client behavior data to obtain normalized client behavior data.
Normalizing bank customer behavior data by using a dispersion normalization method, and performing a correlation on each dimension of characteristics x 1 ,…,x i ,…,x n Where n is the number of clients to be predicted, performObtaining normalized customer behavior data y 1 ,…,y i ,…,y n ∈[0,1]And (5) completing standardization of the customer behavior data.
Step 102: and determining a first objective function according to the characteristic type of the first behavior data, and determining a second objective function according to the characteristic type of the second behavior data.
Converting the bank customer loss prediction problem into a multi-objective optimization problem:
min(f 1 (x)，f 2 (x)),
x＝(a i ,b i ,c i, d i ,e i ,g i ,k i ),a i ∈A,b i ∈B,c i ∈C,d i ∈D,e i ∈E,k i ∈K,g i ∈G,
wherein f 1 (x) Representing a first objective function reflecting customer liveness; f (f) 2 (x) Representing a second objective function reflecting customer satisfaction.
The first objective function is generated by weighting and integrating five characteristic types of estimated annual income, annual average deposit, annual use number, annual average transaction number and number of products owned by the customer, and is as follows:
wherein a is i Representing estimated annual revenue for the ith bank customer, b i Representing the annual average deposit of the ith bank customer, c i, Represents the years of use of the ith bank client, d i Represents the annual average number of transactions of the ith bank customer, e i Indicating the number of owned products of the ith bank customer.
The bank customer satisfaction is obtained by weighting and integrating the online service evaluation data and the offline evaluation data of the customers, and the second objective function is as follows:
in the formula g i On-line service evaluation data, k, representing an ith bank customer i Offline service evaluation data representing an i-th bank customer.
Step 103: a first function value and a second function value for each client to be predicted are determined. The first function value is obtained by inputting first row data into a first objective function; the second function value is obtained by inputting second behavior data into the second objective function.
Step 104: and selecting one client to be predicted from the client to be predicted set as a first target client.
If there are 6 clients to be predicted in the client set to be predicted, client 1, client 2, client 3, client 4, client 5 and client 6 respectively. One of the clients to be predicted is selected as the first target client, e.g. client 1 is selected as the first target client, and steps 105-107 are performed. After the client 1 has finished its prediction, the remaining clients are selected as the first target clients, and steps 105-107 are then performed.
Step 105: and judging whether a second target client exists in the client set to be predicted, wherein the first function value of the second target client is larger than that of the first target client, and the second function value of the second target client is larger than that of the first target client.
Each client to be predicted in the client set to be predicted corresponds to a first function value f 1 And a second function value f 2 . As shown in fig. 2, the function value corresponding to the client 1 is (f 1 (1) ，f 2 (1) ) The function value corresponding to the client 2 is (f) 1 (2) ，f 2 (2) ) The function value corresponding to the client 3 is (f 1 (3) ，f 2 (3) ) GuestThe function value corresponding to user 4 is (f) 1 (4) ，f 2 (4) ) The function value corresponding to the client 5 is (f 1 (5) ，f 2 (5) ) The function value corresponding to the client 6 is (f 1 (6) ，f 2 (6) )。
As can be seen from fig. 2:
if client 1 is considered a first target client, then a second target client is client 4, client 5, and client 6.
If client 2 is considered to be the first target client, then the second target client is client 6.
If client 3 is the first target client, then there is no second target client.
If client 4 is considered to be the first target client, then the second target client is client 6.
If client 5 is the first target client, then there is no second target client.
If client 6 is the first target client, then there is no second target client.
Step 106: if the second target client exists, judging whether a third target client exists in the client set to be predicted, wherein the first function value of the third target client is smaller than the first function value of the first target client, and the second function value of the first target client is smaller than the second function value of the first target client.
The first target clients, where the second target client exists, are client 1, client 2 and client 4.
When the first target client is client 1, there is no third target client.
When the first target client is client 2, there is no third target client.
When the first target client is client 4, there is a third client, client 1.
Step 107: if the third target client does not exist, the first target client is determined to be the first potential attrition client.
If the third target client is not client 1 and client 2, then it is determined that client 1 and client 2 are the first potential attrition clients.
As can be seen from FIG. 2, the present invention utilizes two function values of customer liveness and evaluation degree to obtain two potential loss customers, namely customer 1 and customer 2, respectively, the first function value f of customer 1 1 (1) Is the lowest of all clients and the first function value f of client 2 1 (2) And a second function value f 2 (2) At a lower level in all customers.
As an alternative embodiment, after step 104, the customer churn prediction method further includes:
determining the maximum value and the minimum value of the first function value of the clients to be predicted in the client set to be predicted, and obtaining the maximum first function value and the minimum first function value; determining the maximum value and the minimum value of the second function value of the clients to be predicted in the client set to be predicted, and obtaining the maximum second function value and the minimum second function value;
and determining that the first target client is the first potential churn client when the first function value of the first target client is the maximum first function value and the second function value of the first target client is the minimum second function value, or when the first function value of the first target client is the minimum first function value and the second function value of the first target client is the maximum second function value.
As can be taken from FIG. 2, the maximum first function value is f 1 (3) The minimum first function value is f 1 (1) The maximum second function value is f 2 (5) The minimum second function value is f 2 (3) . Since the first function value of the client 3 is the maximum function value and the second function value of the client 3 is the minimum second function value, the client 3 is determined to be the first potential attrition client. As can be seen from fig. 2, the first function value f of the client 3 1 (3) And a second function value f 2 (3) At a lower level in all customers.
As an alternative embodiment, the customer churn prediction method further includes:
if the second target client does not exist, the first target client is stored in the first target client set; if the third target client exists, the first target client is stored in the first target client set; wherein none of the clients in the first set of target clients are first potential attrition clients;
selecting a first target client from the first target client set as a fourth target client, and judging whether a fifth target client exists in the client set to be predicted, wherein a first function value of the fifth target client is smaller than a first function value of the fourth target client, and a second function value of the fifth target client is smaller than a second function value of the fourth target client;
if the fifth target clients exist, determining the number of the fifth target clients, and if the number of the fifth target clients is 1, determining the fourth target clients as second potential loss clients; the customer churn probability of the second potential churn customer is less than the customer churn probability of the first potential churn customer.
There is no second target customer and not the first potential attrition customer with customers 5 and 6, and there is a third target customer and not the first potential attrition customer with customers 4, 5 and 6 in the first target customer set.
When the fourth target client is client 4, there is a fifth target client that is client 1.
When the fourth target client is client 5, there is a fifth target client that is client 1.
When the fourth target client is client 6, there are fifth target clients that are client 1, client 2, and client 4.
The number of clients of the fifth target clients corresponding to the clients 4 and 5 is 1, and thus the clients 4 and 5 are determined to be the second potential attrition clients. As can be seen from fig. 2, the first function value f of the client 4 1 (4) And a second function value f 2 (4) At a centered level among all customers; second function value f of client 5 2 (5) Although highest among all clients, its second function value f 1 (5) But at a lower level in all customers. Overall, the customer churn probability of the second potential churn customer is less than the customer churn probability of the first potential churn customer.
As an alternative embodiment, after determining that the fourth target client is the second potential attrition client, the method further comprises:
deleting the first target client and the fourth target client in the client to be predicted set, taking the rest clients to be predicted as sixth target clients, wherein the client loss probability of the sixth target clients is smaller than the client loss probability of the second potential loss clients.
Deleting the first target client and the fourth target client in the client set to be predicted, and taking the rest of clients 6 as sixth target clients, wherein as can be seen from fig. 2, the first function value f of the clients 6 1 (6) And a second function value f 2 (6) At a higher level in all customers, customer 6 has a lower churn probability than customers 1, 2, 3, 4, and 5.
According to the method and the system for predicting the loss of the clients, the clients to be predicted are divided from high to low according to the loss probability, and various schemes are formulated for saving the clients according to different loss probability grades (the loss probability is sequentially a first potential loss client, a second potential loss client and a sixth target client from high to low), so that the blindness of saving the clients is solved, and the purpose of stabilizing the clients is achieved.
The present invention also provides a customer churn prediction system, as shown in fig. 3, which includes:
a client data obtaining module 301, configured to obtain a set of clients to be predicted and client behavior data of each client to be predicted in the set of clients to be predicted; the customer behavior data includes first behavior data and second behavior data; the first behavior data reflects the activity degree of the client, and the second behavior data reflects the satisfaction degree of the client;
an objective function determining module 302, configured to determine a first objective function according to the feature type of the first behavior data, and determine a second objective function according to the feature type of the second behavior data;
a function value determining module 303, configured to determine a first function value and a second function value of each client to be predicted; the first function value is obtained by inputting first row data into a first objective function; the second function value is obtained by inputting second behavior data into a second objective function;
the client selection module 304 is configured to select one to-be-predicted client from the to-be-predicted client set as the first target client.
A first determining module 305, configured to determine whether a second target client exists in the set of clients to be predicted, where a first function value of the second target client is greater than a first function value of the first target client and a second function value of the second target client is greater than a second function value of the first target client;
a second judging module 306, configured to judge whether a third target client exists in the to-be-predicted client set when the second target client exists, where a first function value of the third target client is smaller than a first function value of the first target client and a second function value of the first target client is smaller than a second function value of the first target client;
the first customer churn prediction module 307 is configured to determine that the first target customer is the first potential churn customer when the third target customer does not exist.
The customer churn prediction system further comprises:
the second customer loss prediction module is used for determining the maximum value and the minimum value of the first function value of the customers to be predicted in the customer set to be predicted, and obtaining the maximum first function value and the minimum first function value; determining the maximum value and the minimum value of the second function value of the clients to be predicted in the client set to be predicted, and obtaining the maximum second function value and the minimum second function value; and determining that the first target client is the first potential churn client when the first function value of the first target client is the maximum first function value and the second function value of the first target client is the minimum second function value, or when the first function value of the first target client is the minimum first function value and the second function value of the first target client is the maximum second function value.
The third customer loss prediction module is used for storing the first target customers into the first target customer set if the second target customers do not exist or the third target customers exist; the clients in the first set of target clients do not include the first potential attrition client; selecting a first target client from the first target client set as a fourth target client, and judging whether a fifth target client exists in the client set to be predicted, wherein a first function value of the fifth target client is smaller than a first function value of the fourth target client, and a second function value of the fifth target client is smaller than a second function value of the fourth target client; if the fifth target clients exist, determining the number of the fifth target clients, and if the number of the fifth target clients is 1, determining the fourth target clients as second potential loss clients; the customer churn probability of the second potential churn customer is less than the customer churn probability of the first potential churn customer.
And the fourth client loss prediction module is used for deleting the first target client and the fourth target client in the client set to be predicted, taking the rest clients to be predicted as sixth target clients, wherein the client loss probability of the sixth target clients is smaller than the client loss probability of the second potential loss clients.
The customer churn prediction system further comprises:
the unified processing module is used for judging whether the characteristic types of the clients to be predicted in the client set to be predicted are the same after the client behavior data of each client to be predicted in the client set to be predicted is acquired; and if the feature types of the clients to be predicted are different, carrying out unified processing on the feature types of the clients to be predicted.
The abnormality processing module is used for judging whether the client behavior data corresponding to the characteristic type of the client to be predicted in the client set to be predicted is abnormal or not after the client behavior data of the client to be predicted in the client set to be predicted and each client to be predicted in the client set to be predicted are acquired; if so, replacing the abnormal customer behavior data with preset data.
And the normalization processing module is used for carrying out normalization processing on the client behavior data after acquiring the client behavior data of each client to be predicted in the client set to be predicted and each client to be predicted in the client set to be predicted, so as to obtain normalized client behavior data.
An embodiment of the present invention provides a computer-readable storage medium having a program stored thereon, which when executed by a processor, implements the customer churn prediction method described above.
An embodiment of the present invention provides an electronic device, as shown in fig. 4, where the electronic device 40 includes at least one processor 401, and at least one memory 402 and a bus 403 connected to the processor 401; wherein, the processor 401 and the memory 402 complete the communication with each other through the bus 403; the processor 401 is configured to call the program instructions in the memory 402 to perform the customer churn prediction method described above. The electronic device herein may be a server, a PC, a PAD, a mobile phone, etc.
The present application also provides a computer program product adapted to perform a program initialized with the steps comprised by the client churn prediction method described above when executed on a data processing device.
It should be noted that the client loss prediction method, the client loss prediction system, the storage medium and the electronic equipment provided by the invention can be applied to the field of artificial intelligence or the field of finance. The foregoing is merely exemplary, and the application fields of the client churn prediction method, the client churn prediction system, the storage medium and the electronic device provided by the present invention are not limited.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, systems and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, the device includes one or more processors (CPUs), memory, and a bus. The device may also include input/output interfaces, network interfaces, and the like.
The memory may include volatile memory, random Access Memory (RAM), and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip. Memory is an example of a computer-readable medium.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.
In this specification, each embodiment is described in a related manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.
Claims (10)
1. A customer churn prediction method, comprising:
acquiring a client behavior data of each client to be predicted in a client set to be predicted; the customer behavior data comprises first behavior data and second behavior data; the first behavior data reflects the activity degree of the client, and the second behavior data reflects the satisfaction degree of the client;
determining a first objective function according to the characteristic type of the first behavior data, and determining a second objective function according to the characteristic type of the second behavior data;
the first objective function f is generated by weighting and integrating five characteristic types of estimated annual income, annual average deposit, annual use number, annual average transaction number and number of products owned by the customer 1 (x) The method comprises the following steps:
wherein a is i Representing estimated annual revenue for the ith bank customer, b i Representing the annual average deposit of the ith bank customer, c i Represents the years of use of the ith bank client, d i Represents the annual average number of transactions of the ith bank customer, e i Representing the number of owned products of the ith bank customer;
the customer satisfaction degree of the bank is obtained by weighting and integrating the online service evaluation data and the offline evaluation data of the customer, and a second objective function f 2 (x) The method comprises the following steps:
in the formula g i On-line service evaluation data, k, representing an ith bank customer i Offline service evaluation data representing an i-th banking customer;
determining a first function value and a second function value of each client to be predicted; the first function value is obtained by inputting the first row of data into the first objective function; the second function value is obtained by inputting the second behavior data into the second objective function;
selecting one to-be-predicted client from the to-be-predicted client set as a first target client, and judging whether a second target client exists in the to-be-predicted client set, wherein a first function value of the second target client is larger than a first function value of the first target client, and a second function value of the second target client is larger than a second function value of the first target client;
if the second target client exists, judging whether a third target client exists in the client set to be predicted, wherein a first function value of the third target client is smaller than a first function value of the first target client, and a second function value of the first target client is smaller than a second function value of the first target client;
and if the third target client does not exist, determining that the first target client is a first potential attrition client.
2. The customer churn prediction method of claim 1, wherein after selecting one customer to be predicted from the set of customers to be predicted as the first target customer, the method further comprises:
determining the maximum value and the minimum value of the first function values of the clients to be predicted in the client set to be predicted, and obtaining the maximum first function value and the minimum first function value; determining the maximum value and the minimum value of the second function values of the clients to be predicted in the client set to be predicted, and obtaining the maximum second function value and the minimum second function value;
and determining that the first target client is a first potential loss client when the first function value of the first target client is the maximum first function value and the second function value of the first target client is the minimum second function value or when the first function value of the first target client is the minimum first function value and the second function value of the first target client is the maximum second function value.
3. The customer churn prediction method according to claim 1, further comprising:
if the second target client does not exist or the third target client exists, the first target client is stored in a first target client set; the clients in the first set of target clients do not include the first potential attrition client;
selecting a first target client from the first target client set as a fourth target client, and judging whether a fifth target client exists in the client set to be predicted, wherein a first function value of the fifth target client is smaller than a first function value of the fourth target client, and a second function value of the fifth target client is smaller than a second function value of the fourth target client;
if the fifth target client exists, determining the number of the fifth target clients, and if the number of the fifth target clients is 1, determining the fourth target client as a second potential loss client; the customer churn probability of the second potential churn customer is less than the customer churn probability of the first potential churn customer.
4. The customer churn prediction method of claim 3 wherein after said determining said fourth target customer is a second potential churn customer, said method further comprises:
and deleting the first target client and the fourth target client in the client to be predicted set, and taking the rest clients to be predicted as sixth target clients, wherein the client loss probability of the sixth target clients is smaller than the client loss probability of the second potential loss clients.
5. The customer churn prediction method of claim 1, wherein after obtaining the set of customers to be predicted and the customer behavior data of each customer to be predicted in the set of customers to be predicted, the method further comprises:
judging whether the feature types of clients to be predicted in the client set to be predicted are the same or not;
and if the feature types of the clients to be predicted are different, carrying out unified processing on the feature types of the clients to be predicted.
6. The customer churn prediction method of claim 1, wherein after obtaining the set of customers to be predicted and the customer behavior data of each customer to be predicted in the set of customers to be predicted, the method further comprises:
judging whether the client behavior data corresponding to the characteristic type of the client to be predicted in the client set to be predicted is abnormal or not;
if so, replacing the abnormal customer behavior data with preset data.
7. The customer churn prediction method of claim 1, wherein after obtaining the set of customers to be predicted and the customer behavior data of each customer to be predicted in the set of customers to be predicted, the method further comprises:
and carrying out normalization processing on the client behavior data to obtain normalized client behavior data.
8. A customer churn prediction system, comprising:
the client data acquisition module is used for acquiring a client set to be predicted and client behavior data of each client to be predicted in the client set to be predicted; the customer behavior data comprises first behavior data and second behavior data; the first behavior data reflects the activity degree of the client, and the second behavior data reflects the satisfaction degree of the client;
the objective function determining module is used for determining a first objective function according to the characteristic type of the first behavior data and determining a second objective function according to the characteristic type of the second behavior data;
the first objective function f is generated by weighting and integrating five characteristic types of estimated annual income, annual average deposit, annual use number, annual average transaction number and number of products owned by the customer 1 (x) The method comprises the following steps:
wherein a is i Representing estimated annual revenue for the ith bank customer, b i Representing the annual average deposit of the ith bank customer, c i Represents the years of use of the ith bank client, d i Represents the annual average number of transactions of the ith bank customer, e i Representing the number of owned products of the ith bank customer;
the customer satisfaction degree of the bank is obtained by weighting and integrating the online service evaluation data and the offline evaluation data of the customer, and a second objective function f 2 (x) The method comprises the following steps:
in the formula g i On-line service evaluation data, k, representing an ith bank customer i Offline service evaluation data representing an i-th banking customer;
the function value determining module is used for determining a first function value and a second function value of each client to be predicted; the first function value is obtained by inputting the first row of data into the first objective function; the second function value is obtained by inputting the second behavior data into the second objective function;
the first judging module is used for selecting one to-be-predicted client from the to-be-predicted client set as a first target client and judging whether a second target client exists in the to-be-predicted client set, wherein a first function value of the second target client is larger than a first function value of the first target client, and a second function value of the second target client is larger than a second function value of the first target client;
the second judging module is used for judging whether a third target client exists in the client set to be predicted or not when the second target client exists, wherein the first function value of the third target client is smaller than the first function value of the first target client, and the second function value of the first target client is smaller than the second function value of the first target client;
and the first customer churn prediction module is used for determining that the first target customer is a first potential churn customer when the third target customer does not exist.
9. A computer readable storage medium, wherein a program is stored on the computer readable storage medium, which when executed by a processor, implements the customer churn prediction method of any one of claims 1-7.
10. An electronic device, comprising:
at least one processor, and at least one memory, bus, connected to the processor;
the processor and the memory complete communication with each other through the bus; the processor is configured to invoke program instructions in the memory to perform the customer churn prediction method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111068135.8A CN113723710B (en) | 2021-09-13 | 2021-09-13 | Customer loss prediction method, system, storage medium and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111068135.8A CN113723710B (en) | 2021-09-13 | 2021-09-13 | Customer loss prediction method, system, storage medium and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113723710A CN113723710A (en) | 2021-11-30 |
| CN113723710B true CN113723710B (en) | 2024-02-27 |
Family
ID=78683460
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111068135.8A Active CN113723710B (en) | 2021-09-13 | 2021-09-13 | Customer loss prediction method, system, storage medium and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113723710B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889724A (en) * | 2019-11-22 | 2020-03-17 | 北京明略软件系统有限公司 | Customer churn prediction method, customer churn prediction device, electronic equipment and storage medium |
| CN112232945A (en) * | 2020-10-15 | 2021-01-15 | 中国农业银行股份有限公司 | Method and device for determining personal customer credit |
| CN112288117A (en) * | 2019-07-23 | 2021-01-29 | 贝壳技术有限公司 | Target customer deal probability prediction method and device and electronic equipment |
| CN112819258A (en) * | 2021-03-24 | 2021-05-18 | 中国工商银行股份有限公司 | Bank branch to store customer quantity prediction method and device |
-
2021
- 2021-09-13 CN CN202111068135.8A patent/CN113723710B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112288117A (en) * | 2019-07-23 | 2021-01-29 | 贝壳技术有限公司 | Target customer deal probability prediction method and device and electronic equipment |
| CN110889724A (en) * | 2019-11-22 | 2020-03-17 | 北京明略软件系统有限公司 | Customer churn prediction method, customer churn prediction device, electronic equipment and storage medium |
| CN112232945A (en) * | 2020-10-15 | 2021-01-15 | 中国农业银行股份有限公司 | Method and device for determining personal customer credit |
| CN112819258A (en) * | 2021-03-24 | 2021-05-18 | 中国工商银行股份有限公司 | Bank branch to store customer quantity prediction method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113723710A (en) | 2021-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA3065807C (en) | System and method for issuing a loan to a consumer determined to be creditworthy | |
| US10504120B2 (en) | Determining a temporary transaction limit | |
| US20180253657A1 (en) | Real-time credit risk management system | |
| US10614073B2 (en) | System and method for using data incident based modeling and prediction | |
| Lizares et al. | Corporate financial distress: The case of publicly listed firms in an emerging market economy | |
| CN111222994A (en) | Client risk assessment method, device, medium and electronic equipment | |
| CN111708883A (en) | Credit credit limit determination method and device based on machine learning and equipment fingerprint | |
| Gao et al. | A structural‐factor approach to modeling high‐dimensional time series and space‐time data | |
| CN114584601B (en) | User loss identification and intervention method, system, terminal and medium | |
| CN114529400A (en) | Consumption loan preauthorization evaluation method, device and medium | |
| CN113723710B (en) | Customer loss prediction method, system, storage medium and electronic equipment | |
| CN116596657A (en) | Loan risk assessment method and device, storage medium and electronic equipment | |
| CN116630020A (en) | Risk assessment method and device, storage medium and electronic equipment | |
| Balcilar et al. | The predictive power of the term spread on inequality in the United Kingdom: an empirical analysis | |
| Zang | Construction of Mobile Internet Financial Risk Cautioning Framework Based on BP Neural Network | |
| CN113052512A (en) | Risk prediction method and device and electronic equipment | |
| CN118244994B (en) | Historical data storage method and device based on cloud computing | |
| Nasution et al. | Credit Risk Detection in Peer-to-Peer Lending Using CatBoost | |
| Zhang | Research on Stock Market Portfolio Optimization Using Stochastic Matrix Theory and Genetic Algorithm | |
| CN117422545A (en) | Credit risk identification method, apparatus, device and storage medium | |
| CN117952688A (en) | Classification method and device for merchants and electronic equipment | |
| KR101323206B1 (en) | Improved simulation distribution data managing method of credit var calculation system for reduction of data storage | |
| CN116662568A (en) | Method and device for analyzing fund flow direction, storage medium and electronic equipment | |
| KR20140101887A (en) | System for investment analysis and method thereof | |
| CN116720094A (en) | Client information clustering method and device, processor and electronic equipment |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |