CN114219605A - Wind control method, device and storage medium - Google Patents
Wind control method, device and storage medium Download PDFInfo
- Publication number
- CN114219605A CN114219605A CN202111334720.8A CN202111334720A CN114219605A CN 114219605 A CN114219605 A CN 114219605A CN 202111334720 A CN202111334720 A CN 202111334720A CN 114219605 A CN114219605 A CN 114219605A
- Authority
- CN
- China
- Prior art keywords
- risk rule
- initial
- pools
- risk
- pool
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Physics & Mathematics (AREA)
- Technology Law (AREA)
- Biomedical Technology (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Artificial Intelligence (AREA)
- General Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present disclosure provides a wind control method and device, the method comprising: coding the risk rule to be selected; dividing all the risk rules to be selected to form at least one initial risk rule pool, wherein each initial risk rule pool comprises at least one risk rule to be selected, and all the initial risk rule pools form an initial population; calculating KS values of the initial risk rule pools; and performing population evolution of a preset generation number on the initial population based on the coding values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool, and performing pre-credit risk prediction on the client based on the optimal risk rule pool so as to perform wind control. The method provided by the disclosure can enable the finally constructed risk rule pool to identify risks more accurately, improves the accuracy of risk judgment, and has higher efficiency.
Description
Technical Field
The disclosure relates to the technical field of big data intelligent analysis, in particular to a wind control method, a wind control device and a storage medium.
Background
In recent years, the development of the general financial loan service is rapid, and the market competition in the field is more and more intense. Under the background, the bank needs to continuously innovate from the aspects of improving customer experience, handling efficiency, accurately controlling risks and the like, so that product homogenization is avoided, product competitiveness is improved, and the favor of customers and a cooperation platform is obtained. Meanwhile, the pre-loan wind control requirement of a popular financial loan scene is continuously improved, and the requirements on the aspects of regional characteristic risk identification, fine wind control management and the like are increasingly strict, so that higher requirements are provided for the construction of a risk rule pool.
In the related art, a strong rule method is usually adopted when the risk rule pool is constructed. However, the strong rule method mainly depends on the experience of the wind control professional, and the wind control professional designs the wind control professional based on the past product experience, so that the labor cost is high. Meanwhile, when the number of indexes to be selected is large, the function of the weak rule cannot be reflected only by manual selection, so that the risk identification capability of the constructed risk rule pool is low, and meanwhile, the updating iteration is slow.
Disclosure of Invention
The present disclosure provides a wind control method and device, which are used for solving the technical problems of high labor cost, low risk identification capability and slow update iteration when a risk rule pool is constructed in the related art.
An embodiment of a first aspect of the present disclosure provides a wind control method, including:
coding the risk rule to be selected;
dividing all the risk rules to be selected to form at least one initial risk rule pool, wherein each initial risk rule pool comprises at least one risk rule to be selected, and all the initial risk rule pools form an initial population;
calculating KS values of the initial risk rule pools;
and performing population evolution of a preset generation number on the initial population based on the coding values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool, and performing pre-credit risk prediction on the client based on the optimal risk rule pool so as to perform wind control.
An embodiment of a second aspect of the present disclosure provides a wind control device, where the system includes:
the coding module is used for coding the risk rule to be selected;
the system comprises a dividing module, a selecting module and a selecting module, wherein the dividing module is used for dividing all risk rules to be selected to form at least one initial risk rule pool, each initial risk rule pool comprises at least one risk rule to be selected, and all initial risk rule pools form an initial population;
the calculation module is used for calculating KS values of the initial risk rule pools;
and the evolution module is used for performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool, and performing pre-credit risk prediction on the client based on the optimal risk rule pool so as to perform wind control.
A third embodiment of the present disclosure provides a computer storage medium, where the computer storage medium stores computer-executable instructions; the computer-executable instructions, when executed by a processor, enable the method as described above.
A computer device according to an embodiment of a fourth aspect of the present disclosure includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the method according to the first aspect is implemented.
The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:
to sum up, in the wind control method and the wind control device provided in the embodiments of the present disclosure, the risk rules to be selected are encoded, and all the risk rules to be selected are divided to form at least one initial risk rule pool, where all the initial risk rule pools form an initial population; then, calculating KS values of the initial risk rule pools; and performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, and determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool. Therefore, in the embodiment of the disclosure, through the application of the genetic algorithm, the initial population processed from a large amount of data is subjected to continuous population evolution, and the risk rule pool is finally constructed, so that the application of the weak rule in the risk rule pool can be explored, the finally constructed risk rule pool can identify risks more accurately, and the accuracy of risk judgment is improved.
Meanwhile, when a new sample or a new data dimension appears, only the initial population needs to be adjusted correspondingly, population evolution can be carried out again based on the genetic algorithm to quickly establish a risk rule pool, and the efficiency is high.
In addition, in the embodiment of the present disclosure, in the genetic algorithm, the calculation of the KS value is also used as a fitness function, so that the accuracy of constructing the risk rule pool can be further improved.
Additional aspects and advantages of the disclosure will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the disclosure.
Drawings
The foregoing and/or additional aspects and advantages of the present disclosure will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a schematic flow chart of a wind control method according to an embodiment of the present disclosure;
fig. 2 is a schematic flowchart of a population evolution method for performing a predetermined algebra on an initial population according to a second embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of a wind control device according to a third embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of an evolution module according to a fourth embodiment of the present disclosure.
Detailed Description
Reference will now be made in detail to the embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and intended to be illustrative of the present disclosure, and should not be construed as limiting the present disclosure.
The following explains terms to which the present disclosure relates.
Risk rule pool: the risk admission model is composed of a plurality of rules, each rule in the pool has a corresponding score, and the final score of a client is the sum of the corresponding scores of all the rules hit by the client.
KS (Kolmogorov-Smirnov) value: the method is an index for evaluating the distinguishing capability of the model, the results (generally scores) of the model are segmented and sequenced, the difference value of the good account ratio and the bad account ratio in each score interval is calculated, and the maximum value of the difference value is the KS value.
The following describes a wind control method and system of an embodiment of the present disclosure with reference to the drawings.
Example one
Fig. 1 is a schematic flow chart of a wind control method according to an embodiment of the present disclosure, and as shown in fig. 1, the method may include:
Each candidate risk rule may be binary-coded, so that each candidate risk rule corresponds to a unique coding value, where the coding value may be referred to as a gene.
For example, when there are 40 risk rule rules to be selected, since 25＜40＜26Thus, a 6-bit binary number may be used for encoding, e.g., so that the first risk rule corresponds to the code 000001, the second risk rule corresponds to the code 000010, the third risk rule corresponds to the code 000011, and so on.
And 102, dividing all the risk rules to be selected to form at least one initial risk rule pool, wherein each initial risk rule pool comprises at least one risk rule to be selected, and all the initial risk rule pools form an initial population.
Specifically, a plurality of initial risk rule pools may be randomly generated, and each initial risk rule pool includes a plurality of risk rules to be selected. And each initial pool of risk rules may be referred to as an initial individual, with all initial individuals comprising an initial population.
All risk rules contained in the initial risk rule pool are called as chromosomes, and the number of the risk rules contained in each initial risk rule pool can be controlled by specifying the length of the chromosomes; for example: when it is desired to control the number of risk rules in the initial risk rule pool to be 10, and one risk rule is coded by a 6-bit binary number, it is only necessary to specify a chromosome length of 10 × 6 — 60 bits.
And 103, calculating the KS value of each initial risk rule pool.
The risk rules included in each initial risk rule pool may be determined based on the chromosomes corresponding to each initial risk rule pool, and then, the KS values of each risk rule pool are calculated based on the risk rules included in the initial risk rule pool, and the KS values are used as the fitness function of the initial risk rule pool.
And 104, performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, and determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool.
After the optimal risk rule pool is determined, pre-loan risk prediction can be performed on the client based on the optimal risk rule pool so as to perform wind control.
To sum up, in the wind control method provided by the embodiment of the present disclosure, the risk rules to be selected are encoded, and all the risk rules to be selected are divided to form at least one initial risk rule pool, where all the initial risk rule pools form an initial population; then, calculating KS values of the initial risk rule pools; and performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, and determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool. Therefore, in the embodiment of the disclosure, through the application of the genetic algorithm, the initial population processed from a large amount of data is subjected to continuous population evolution, and the risk rule pool is finally constructed, so that the application of the weak rule in the risk rule pool can be explored, the finally constructed risk rule pool can identify risks more accurately, and the accuracy of risk judgment is improved.
Meanwhile, when a new sample or a new data dimension appears, only the initial population needs to be adjusted correspondingly, population evolution can be carried out again based on the genetic algorithm to quickly establish a risk rule pool, and the efficiency is high.
In addition, in the embodiment of the present disclosure, in the genetic algorithm, the calculation of the KS value is also used as a fitness function, so that the accuracy of constructing the risk rule pool can be further improved.
Example two
The technique used in step 104 above is a genetic algorithm. The genetic algorithm is a calculation model of a biological evolution process for simulating natural selection and genetic mechanism of Darwinian biological evolution theory, and is a method for searching an optimal solution by simulating the natural evolution process. The method is mainly characterized in that the method directly operates the structural object without the limitation of derivation and function continuity; the method has the advantages of inherent hidden parallelism and better global optimization capability; by adopting a probabilistic optimization method, the optimized search space can be automatically acquired and guided without a determined rule, and the search direction can be adaptively adjusted. Genetic algorithms target all individuals in a population and use randomization techniques to guide an efficient search of an encoded parameter space. Wherein the selection, crossover and mutation constitute genetic operations of the genetic algorithm; the core content of the genetic algorithm is composed of five elements of parameter coding, initial population setting, fitness function design, genetic operation design and control parameter setting.
The genetic algorithm simulates the evolution process of an artificial population, a group of candidate individuals are reserved in each iteration through mechanisms such as selection, intersection, variation and the like, the process is repeated, and after the population is evolved for a plurality of generations, the fitness of the population reaches an approximately optimal state under an ideal condition.
Specifically, fig. 2 is a schematic flow chart of a method for performing population evolution of a preset generation number on an initial population according to a second embodiment of the present disclosure, and as shown in fig. 2, the performing population evolution of the preset generation number on the initial population based on the code value and the KS value corresponding to each initial risk rule pool in step 104 by using a genetic algorithm may specifically include:
and 1041, sequencing the initial risk rule pools according to the sequence of KS values from high to low, and reserving the first N initial risk rule pools, wherein N is a positive integer and is less than or equal to the total number of the initial risk rule pools included in the initial population.
After the initial risk rule pool is sorted from high to low KS values, different selection probabilities may be assigned to different KS values, and the higher the KS value is, the higher the assigned selection probability is. For example, one can specify that the initial risk rule pool with the highest KS value is selected with a probability of 95%, the initial risk rule pool with the second highest KS value is selected with a probability of 90%, and so on.
1042, performing pairwise crossing processing on the first N initial risk rule pools.
The interleaving process comprises: and selecting two initial risk rule pools from the first N initial risk rule pools, crossing the coding values corresponding to the two initial risk rule pools by adopting a single-point crossing method to obtain two new risk rule pools, calculating KS values of the two new risk rule pools, and reserving a risk rule pool with a higher KS value in the two new risk rule pools.
It should be noted that, the code values corresponding to the initial risk rule pool mainly include: consisting of the coded values of all risk rules in the initial risk rule pool in their rank order. For example, assume that the initial risk rule a includes two risk rules, namely risk rule 1 and risk rule 2, where the risk rule 1 is arranged at the front, the risk rule 2 is arranged at the back, and the coding value corresponding to the risk rule 1 is a1The risk rule 2 corresponds to a code value of a2Then, it can be determined that the code value corresponding to the initial risk rule pool is a1a2。
And the step of crossing the code values corresponding to the two initial risk rule pools by using the single-point crossing method to obtain two new risk rule pools mainly comprises the following steps:
cutting the coding value of each initial risk rule pool of the two initial risk rule pools at a coding point position to obtain four segments of sub-codes; and combining the four sections of sub-codes two by two again to obtain two sections of new codes, and recombining the risk rules corresponding to each section of new codes to form a new risk rule pool.
By way of example, assume that the initial risk rule pool A has an encoded value of a1a2The code of the initial risk rule pool B is B1b2Then the encoded value a of the initial risk rule pool A may be set1a2Is divided into1And a2Encoding value B of initial risk rule pool B1b2Is divided into1And b2Then a is added1、a2、b1、b2Recombined into two new segments of code a1b2And b1a2. A is to1b2Corresponding risk rules are recombined into a new risk rule pool, b1a2And the corresponding risk rules are recombined into a new risk rule pool, and the risk rule pool with the higher KS value in the two new risk rule pools is reserved.
The performing variation processing on the code value corresponding to the cross-processed risk rule pool by using a single-point variation method specifically includes the following steps:
step a, randomly selecting a section of sub-codes in the code values of the risk rule pool after the cross processing, and randomly generating a variation seed, wherein the bit number of the variation seed is the same as that of the sub-codes.
For example, assuming that the coded value of the risk rule pool after the cross-processing is 000001000010, the first 4 codes in the coded value 000001000010 are regarded as the sub-codes, and the sub-codes are 0000. And randomly generated to mutate the seed 1010.
And b, carrying out bitwise negation or bitwise XOR operation on the sub-codes and the variation seeds to form varied codes.
Assuming that the bitwise xor operation is performed in the present disclosure, the variant code obtained after bitwise xor operation is performed on the sub-code 0000 and the variant seed 1010 is: 1010.
and c, replacing the sub-codes with the mutated codes to obtain a mutated risk rule pool.
The mutated code 1010 is used to replace the original sub-code 0000 to obtain the mutated risk rule pool 101001000010.
And step 1044, sequencing the risk rule pools subjected to the mutation treatment according to the sequence from high KS value to low KS value, and reserving the first M initial risk rule pools to form a new generation population, wherein M is a positive integer and is not more than N.
Performing the above steps 1041-1044 can realize generation population evolution on the risk rule pool, and repeating the above steps 1041-1044 until a preset generation number (for example, 20 or 100), determining the preset generation population as the last generation population, and determining the risk rule pool with the highest KS value in the last generation population as the optimal risk rule pool.
To sum up, in the wind control method provided by the embodiment of the present disclosure, the risk rules to be selected are encoded, and all the risk rules to be selected are divided to form at least one initial risk rule pool, where all the initial risk rule pools form an initial population; then, calculating KS values of the initial risk rule pools; and performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, and determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool. Therefore, in the embodiment of the disclosure, through the application of the genetic algorithm, the initial population processed from a large amount of data is subjected to continuous population evolution, and the risk rule pool is finally constructed, so that the application of the weak rule in the risk rule pool can be explored, the finally constructed risk rule pool can identify risks more accurately, and the accuracy of risk judgment is improved.
Meanwhile, when a new sample or a new data dimension appears, only the initial population needs to be adjusted correspondingly, population evolution can be carried out again based on the genetic algorithm to quickly establish a risk rule pool, and the efficiency is high.
In addition, in the embodiment of the present disclosure, in the genetic algorithm, the calculation of the KS value is also used as a fitness function, so that the accuracy of constructing the risk rule pool can be further improved.
EXAMPLE III
Fig. 3 is a schematic structural diagram of a wind control device 300 according to a third embodiment of the present disclosure, and as shown in fig. 3, the device may include:
the encoding module 301 is configured to encode the risk rule to be selected;
a dividing module 302, configured to divide all to-be-selected risk rules to form at least one initial risk rule pool, where each initial risk rule pool includes at least one to-be-selected risk rule, and all initial risk rule pools constitute an initial population;
a calculating module 303, configured to calculate a KS value of each initial risk rule pool;
and the evolution module 304 is used for performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool, and performing pre-credit risk prediction on the client based on the optimal risk rule pool so as to perform wind control.
Optionally, the encoding module is further configured to:
each candidate risk rule is corresponding to a unique code value.
Optionally, the encoding is binary encoding.
To sum up, in the wind control device provided in the embodiment of the present disclosure, the risk rules to be selected are encoded, and all the risk rules to be selected are divided to form at least one initial risk rule pool, where all the initial risk rule pools form an initial population; then, calculating KS values of the initial risk rule pools; and performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, and determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool. Therefore, in the embodiment of the disclosure, through the application of the genetic algorithm, the initial population processed from a large amount of data is subjected to continuous population evolution, and the risk rule pool is finally constructed, so that the application of the weak rule in the risk rule pool can be explored, the finally constructed risk rule pool can identify risks more accurately, and the accuracy of risk judgment is improved.
Meanwhile, when a new sample or a new data dimension appears, only the initial population needs to be adjusted correspondingly, population evolution can be carried out again based on the genetic algorithm to quickly establish a risk rule pool, and the efficiency is high.
In addition, in the embodiment of the present disclosure, in the genetic algorithm, the calculation of the KS value is also used as a fitness function, so that the accuracy of constructing the risk rule pool can be further improved.
Example four
Fig. 4 is a schematic structural diagram of an evolution module 304 according to a fourth embodiment of the present disclosure, and as shown in fig. 4, the apparatus may include:
optionally, the evolution module 304 includes:
a first sorting module 3041, configured to sort the initial risk rule pools according to a sequence from high to low of KS values, and keep the first N initial risk rule pools, where N is a positive integer and is less than or equal to the total number of the initial risk rule pools included in the initial population;
an intersection processing module 3042, configured to perform pairwise intersection processing on the first N initial risk rule pools, where the intersection processing includes: optionally selecting two initial risk rule pools from the first N initial risk rule pools, crossing the coding values corresponding to the two initial risk rule pools by adopting a single-point crossing method to obtain two new risk rule pools, calculating KS values of the two new risk rule pools, and reserving a risk rule pool with a higher KS value in the two new risk rule pools;
a mutation processing module 3041, configured to perform mutation processing on the risk rule pool after the cross processing, where the mutation processing includes: carrying out mutation processing on the coding value corresponding to the cross-processed risk rule pool by adopting a single-point mutation method;
a second sorting module 3041, configured to sort the risk rule pools after the mutation processing according to the sequence from high KS value to low KS value, and reserve the first M initial risk rule pools to form a new generation population, where M is a positive integer and is not greater than N;
and repeating the steps until a preset generation number is reached, and determining the preset generation population as the last generation population.
Optionally, the cross processing module is further configured to:
cutting the coding value of each initial risk rule pool of the two initial risk rule pools at a coding point position to obtain four segments of sub-codes; and combining the four sections of sub-codes two by two again to obtain two sections of new codes, and recombining the risk rules corresponding to each section of new codes to form a new risk rule pool.
Optionally, the mutation processing module is further configured to:
randomly selecting a section of sub-codes from the code values of the risk rule pool after the cross processing, and randomly generating a variation seed, wherein the bit number of the variation seed is the same as that of the sub-codes;
carrying out bitwise negation or bitwise XOR operation on the sub-codes and the variant seeds to form variant codes;
and replacing the sub-codes with the mutated codes to obtain a mutated risk rule pool.
Optionally, the preset generation number is 20 or 100 generations
To sum up, in the wind control device provided in the embodiment of the present disclosure, the risk rules to be selected are encoded, and all the risk rules to be selected are divided to form at least one initial risk rule pool, where all the initial risk rule pools form an initial population; then, calculating KS values of the initial risk rule pools; and performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, and determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool. Therefore, in the embodiment of the disclosure, through the application of the genetic algorithm, the initial population processed from a large amount of data is subjected to continuous population evolution, and the risk rule pool is finally constructed, so that the application of the weak rule in the risk rule pool can be explored, the finally constructed risk rule pool can identify risks more accurately, and the accuracy of risk judgment is improved.
Meanwhile, when a new sample or a new data dimension appears, only the initial population needs to be adjusted correspondingly, population evolution can be carried out again based on the genetic algorithm to quickly establish a risk rule pool, and the efficiency is high.
In addition, in the embodiment of the present disclosure, in the genetic algorithm, the calculation of the KS value is also used as a fitness function, so that the accuracy of constructing the risk rule pool can be further improved.
In order to implement the above embodiments, the present disclosure also provides a computer storage medium.
The computer storage medium provided by the embodiment of the disclosure stores an executable program; the executable program, when executed by a processor, is capable of implementing the method as shown in fig. 1 or fig. 2.
The present disclosure also provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the method shown in fig. 1 or fig. 2
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present disclosure in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present disclosure.
Although embodiments of the present disclosure have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure, and that changes, modifications, substitutions and alterations may be made to the above embodiments by those of ordinary skill in the art within the scope of the present disclosure.
Claims (16)
1. A method of wind control, the method comprising:
coding the risk rule to be selected;
dividing all the risk rules to be selected to form at least one initial risk rule pool, wherein each initial risk rule pool comprises at least one risk rule to be selected, and all the initial risk rule pools form an initial population;
calculating KS values of the initial risk rule pools;
and performing population evolution of a preset generation number on the initial population based on the coding values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool, and performing pre-credit risk prediction on the client based on the optimal risk rule pool so as to perform wind control.
2. The wind control method according to claim 1, wherein the encoding the candidate risk rules comprises:
each candidate risk rule is corresponding to a unique code value.
3. The wind control method of claim 2, wherein the code is a binary code.
4. The method of claim 1, wherein performing population evolution with a predetermined algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm comprises:
sequencing the initial risk rule pools according to the KS values from high to low, and reserving the first N initial risk rule pools, wherein N is a positive integer and is less than or equal to the total number of the initial risk rule pools included by the initial population;
performing pairwise crossing processing on the first N initial risk rule pools, wherein the crossing processing comprises the following steps: optionally selecting two initial risk rule pools from the first N initial risk rule pools, crossing the coding values corresponding to the two initial risk rule pools by adopting a single-point crossing method to obtain two new risk rule pools, calculating KS values of the two new risk rule pools, and reserving a risk rule pool with a higher KS value in the two new risk rule pools;
performing mutation processing on the risk rule pool after the cross processing, wherein the mutation processing comprises the following steps: carrying out mutation processing on the coding value corresponding to the cross-processed risk rule pool by adopting a single-point mutation method;
sorting the risk rule pools subjected to the mutation treatment according to the sequence of KS values from high to low, and reserving the first M initial risk rule pools to form a new generation population, wherein M is a positive integer and is less than or equal to N;
and repeating the steps until a preset generation number is reached, and determining the preset generation population as the last generation population.
5. The method according to claim 4, wherein the crossing the encoded values corresponding to the two initial risk rule pools by using a single-point crossing method to obtain two new risk rule pools comprises:
cutting the coding value of each initial risk rule pool of the two initial risk rule pools at a coding point position to obtain four segments of sub-codes; and combining the four sections of sub-codes two by two again to obtain two sections of new codes, and recombining the risk rules corresponding to each section of new codes to form a new risk rule pool.
6. The wind control method according to claim 4, wherein the performing mutation processing on the code value corresponding to the cross-processed risk rule pool by using a single-point mutation method includes:
randomly selecting a section of sub-codes from the code values of the risk rule pool after the cross processing, and randomly generating a variation seed, wherein the bit number of the variation seed is the same as that of the sub-codes;
carrying out bitwise negation or bitwise XOR operation on the sub-codes and the variant seeds to form variant codes;
and replacing the sub-codes with the mutated codes to obtain a mutated risk rule pool.
7. The wind control method according to claim 4, wherein the predetermined generation number is 20 or 100.
8. A wind control device, characterized in that the device comprises:
the coding module is used for coding the risk rule to be selected;
the system comprises a dividing module, a selecting module and a selecting module, wherein the dividing module is used for dividing all risk rules to be selected to form at least one initial risk rule pool, each initial risk rule pool comprises at least one risk rule to be selected, and all initial risk rule pools form an initial population;
the calculation module is used for calculating KS values of the initial risk rule pools;
and the evolution module is used for performing population evolution of a preset algebra on the initial population based on the code values and KS values corresponding to the initial risk rule pools by using a genetic algorithm, determining the risk rule pool with the highest KS value in the population of the last generation as an optimal risk rule pool, and performing pre-credit risk prediction on the client based on the optimal risk rule pool so as to perform wind control.
9. The wind control apparatus of claim 8, wherein the encoding module is further configured to:
each candidate risk rule is corresponding to a unique code value.
10. Wind control device according to claim 9, characterized in that the coding is a binary coding.
11. The wind control apparatus of claim 8, wherein the evolution module is further configured to:
sequencing the initial risk rule pools according to the KS values from high to low, and reserving the first N initial risk rule pools, wherein N is a positive integer and is less than or equal to the total number of the initial risk rule pools included by the initial population;
performing pairwise crossing processing on the first N initial risk rule pools, wherein the crossing processing comprises the following steps: optionally selecting two initial risk rule pools from the first N initial risk rule pools, crossing the coding values corresponding to the two initial risk rule pools by adopting a single-point crossing method to obtain two new risk rule pools, calculating KS values of the two new risk rule pools, and reserving a risk rule pool with a higher KS value in the two new risk rule pools;
performing mutation processing on the risk rule pool after the cross processing, wherein the mutation processing comprises the following steps: carrying out mutation processing on the coding value corresponding to the cross-processed risk rule pool by adopting a single-point mutation method;
sorting the risk rule pools subjected to the mutation treatment according to the sequence of KS values from high to low, and reserving the first M initial risk rule pools to form a new generation population, wherein M is a positive integer and is less than or equal to N;
and repeating the steps until a preset generation number is reached, and determining the preset generation population as the last generation population.
12. The wind control apparatus of claim 11, wherein the evolution module is further configured to:
cutting the coding value of each initial risk rule pool of the two initial risk rule pools at a coding point position to obtain four segments of sub-codes; and combining the four sections of sub-codes two by two again to obtain two sections of new codes, and recombining the risk rules corresponding to each section of new codes to form a new risk rule pool.
13. The wind control apparatus of claim 11, wherein the evolution module is further configured to:
randomly selecting a section of sub-codes from the code values of the risk rule pool after the cross processing, and randomly generating a variation seed, wherein the bit number of the variation seed is the same as that of the sub-codes;
carrying out bitwise negation or bitwise XOR operation on the sub-codes and the variant seeds to form variant codes;
and replacing the sub-codes with the mutated codes to obtain a mutated risk rule pool.
14. The wind control device of claim 11, wherein the predetermined number of generations is 20 or 100 generations.
15. A computer storage medium, wherein the computer storage medium stores computer-executable instructions; the computer-executable instructions, when executed by a processor, are capable of performing the method of any one of claims 1-9.
16. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the program implementing the method according to any of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111334720.8A CN114219605A (en) | 2021-11-11 | 2021-11-11 | Wind control method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111334720.8A CN114219605A (en) | 2021-11-11 | 2021-11-11 | Wind control method, device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114219605A true CN114219605A (en) | 2022-03-22 |
Family
ID=80696975
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111334720.8A Pending CN114219605A (en) | 2021-11-11 | 2021-11-11 | Wind control method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114219605A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115578191A (en) * | 2022-09-21 | 2023-01-06 | 福建润楼金融产业投资有限公司 | Genetic algorithm-based adaptive wind control strategy optimization method |
-
2021
- 2021-11-11 CN CN202111334720.8A patent/CN114219605A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115578191A (en) * | 2022-09-21 | 2023-01-06 | 福建润楼金融产业投资有限公司 | Genetic algorithm-based adaptive wind control strategy optimization method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Elegbede et al. | Availability allocation to repairable systems with genetic algorithms: a multi-objective formulation | |
| Abraham et al. | Evolutionary computation: from genetic algorithms to genetic programming | |
| CN109448794B (en) | Genetic taboo and Bayesian network-based epistatic site mining method | |
| CN111832101A (en) | Construction method of cement strength prediction model and cement strength prediction method | |
| CN112580259B (en) | Intelligent mine automatic ore blending method and system based on genetic algorithm | |
| CN108776461A (en) | A kind of flexible job shop scheduling method and system | |
| Busetti | Genetic algorithms overview | |
| CN112734051A (en) | Evolutionary ensemble learning method for classification problem | |
| CN113128655A (en) | Multi-population genetic algorithm-based industrial control intrusion detection classifier parameter selection method | |
| CN114219605A (en) | Wind control method, device and storage medium | |
| Malik | A study of genetic algorithm and crossover techniques | |
| CN115481727A (en) | Intention recognition neural network generation and optimization method based on evolutionary computation | |
| CN114241267A (en) | Structural entropy sampling-based multi-target architecture search osteoporosis image identification method | |
| CN116305939A (en) | High-precision inversion method and system for carbon water flux of land ecological system and electronic equipment | |
| CN116611504A (en) | Neural architecture searching method based on evolution | |
| CN109615074B (en) | Method and device for generating monster configuration of game | |
| Sivaraj et al. | An efficient grouping genetic algorithm | |
| Bashir et al. | Solving Banana (Rosenbrock) Function based on fitness function | |
| Eiben et al. | Evolutionary algorithms | |
| Lefort et al. | Simultaneous optimization of weights and structure of an RBF neural network | |
| CN113838526B (en) | Virus mutant generation method, system, computer equipment and medium | |
| Mohebpour et al. | Ccgdc: A new crossover operator for genetic data clustering | |
| Chen et al. | Context quantization based on the modified genetic algorithm with K-means | |
| CN111883208B (en) | Gene sequence optimization method, device, equipment and medium | |
| Borana | Genetic algorithms and its application to economic load dispatch |
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 |