Medical insurance fraud detection method and device based on particle swarm optimization generative adversarial network

Abstract

The present invention provides a medical insurance fraud detection method and device based on a particle swarm optimization generative adversarial network, relating to the field of data processing technology. The detection method comprises: inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data; filtering the synthetic sample data using a first particle swarm optimization algorithm to obtain multiple filtered synthetic sample data, wherein each particle in the first particle swarm optimization algorithm corresponds to a sample set including multiple synthetic sample data; combining the multiple filtered synthetic sample data with multiple real sample data to obtain a training sample set; training a classification model based on the training sample set to obtain a medical insurance fraud detection model; and inputting the data to be detected into the medical insurance fraud detection model to obtain a fraud detection result output by the medical insurance fraud detection model. The present invention can improve the accuracy of medical insurance fraud detection.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a medical insurance fraud detection method and device based on a particle swarm optimization-based generative adversarial network. Background Art

[0002] Existing technologies use traditional supervised models (such as logistic regression and random forests) to detect medical insurance fraud. However, due to the extremely low proportion of real-world examples of medical insurance fraud, medical insurance data suffers from extreme class imbalance. During training, the loss function of traditional supervised models is dominated by the majority class, significantly reducing sensitivity to the minority class. This leads to high missed detection rates and poor generalization performance. Existing supervised models for detecting medical insurance fraud suffer from low accuracy. Summary of the Invention

[0003] The present invention provides a medical insurance fraud detection method and device based on a particle swarm optimization generative adversarial network, which is used to solve the defect of low accuracy of medical insurance fraud detection in the prior art and achieve the effect of improving the accuracy of medical insurance fraud detection.

[0004] The present invention provides a medical insurance fraud detection method based on a particle swarm optimization generative adversarial network, comprising:

[0005] Inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, wherein the synthetic sample data includes synthetic sample data to be detected and a fraud label corresponding to the synthetic sample data to be detected, the generative adversarial network also includes a discriminator, and the training data of the generative adversarial network includes real sample data, the real sample data includes real sample data to be detected and a fraud label corresponding to the real sample data to be detected;

[0006] Based on a first particle swarm optimization algorithm, the synthetic sample data is screened to obtain a plurality of screened synthetic sample data, wherein in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set includes a plurality of the synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set;

[0007] combining the plurality of screened synthetic sample data with the plurality of real sample data to obtain a training sample set, and training a classification model based on the training sample set to obtain a medical insurance fraud detection model;

[0008] The data to be detected is input into the medical insurance fraud detection model to obtain the fraud detection result output by the medical insurance fraud detection model.

[0009] According to a medical insurance fraud detection method based on a particle swarm optimization generative adversarial network provided by the present invention, before inputting random noise into the generator of the trained generative adversarial network, the method includes:

[0010] Acquiring real medical insurance data, where the real medical insurance data includes multiple types of features;

[0011] Determining a target feature type from the multiple types of features using a second particle swarm optimization algorithm, wherein a particle in the second particle swarm optimization algorithm corresponds to a feature type set, the feature type set includes at least one feature type, and a fitness value of the particle in the second particle swarm optimization algorithm reflects the discrimination ability and independence of the feature type in the feature type set corresponding to the particle;

[0012] The real sample data is constructed based on the target feature type.

[0013] According to a medical insurance fraud detection method based on a particle swarm optimization generative adversarial network provided by the present invention, the fitness value of the particles in the second particle swarm optimization algorithm is determined based on the first formula;

[0014] The first formula is:

[0015] ;

[0016] in, Indicates the fitness value of the particle whose corresponding feature type set is X, 、 、 is the weight coefficient, represents the mutual information between X and the fraud label set Y, , n is the total number of samples, Represents the value vector of the feature type of the i-th sample in the feature type set X, is the fraud label of the i-th sample, Represents the average value vector of the feature type of all samples in the feature type set X, represents the mean fraud label of all samples, represents the number of feature types in the feature type set X, and are the standard deviation vectors of X and Y respectively.

[0017] According to a medical insurance fraud detection method based on a particle swarm optimization generative adversarial network provided by the present invention, the local density calculation formula in the fitness value of particles in the first particle swarm optimization algorithm is:

[0018] ;

[0019] ;

[0020] in, is the local density of the synthetic sample data s, is the estimated probability density of the synthetic sample data s, k is the total number of the synthetic sample data, Indicates that the synthetic sample data s and the synthetic sample data The Euclidean distance between is the standard deviation of the real sample data; d is the feature dimension in the synthetic sample data, is a normalizing constant.

[0021] According to a medical insurance fraud detection method based on a particle swarm optimization generative adversarial network provided by the present invention, the global diversity calculation formula in the fitness value of particles in the first particle swarm optimization algorithm is:

[0022] ;

[0023] in, represents the global diversity of the sample set corresponding to the particle, Represents the sample set The 1-Wasserstein distance between the distribution of and the distribution of the real sample data.

[0024] According to a medical insurance fraud detection method based on particle swarm optimization generative adversarial network provided by the present invention, the number of synthetic sample data is N times the difference between the number of majority class samples and the number of minority class samples in the real sample data, and N is greater than or equal to 2.

[0025] The present invention also provides a medical insurance fraud detection device based on a particle swarm optimization generative adversarial network, comprising:

[0026] A sample synthesis module is configured to input random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, wherein the synthetic sample data includes synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected, the generative adversarial network also includes a discriminator, and the training data of the generative adversarial network includes real sample data, the real sample data includes real sample data to be detected and fraud labels corresponding to the real sample data to be detected;

[0027] a sample screening module, configured to screen the synthetic sample data based on a first particle swarm optimization algorithm to obtain a plurality of screened synthetic sample data, wherein in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set including a plurality of synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set;

[0028] a training module, configured to combine the plurality of screened synthetic sample data with the plurality of real sample data to obtain a training sample set, and train a classification model based on the training sample set to obtain a medical insurance fraud detection model;

[0029] The detection module is used to input the data to be detected into the medical insurance fraud detection model to obtain the fraud detection result output by the medical insurance fraud detection model.

[0030] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements any of the above-described medical insurance fraud detection methods based on particle swarm optimization generative adversarial networks.

[0031] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for detecting medical insurance fraud based on a particle swarm optimization-based generative adversarial network as described above is implemented.

[0032] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned medical insurance fraud detection methods based on particle swarm optimization generative adversarial networks.

[0033] The present invention provides a medical insurance fraud detection method and device based on a particle swarm optimization generative adversarial network, wherein the medical insurance fraud detection method based on a particle swarm optimization generative adversarial network includes: inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, the synthetic sample data including synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected, the generative adversarial network also including a discriminator, the training data of the generative adversarial network including real sample data, the real sample data including real sample data to be detected and fraud labels corresponding to the real sample data to be detected; based on a first particle swarm optimization algorithm, the synthetic sample data is screened to obtain multiple screened synthetic sample data, in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set includes multiple synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set; the multiple screened synthetic sample data and the multiple real sample data are combined to obtain a training sample set, a classification model is trained based on the training sample set to obtain a medical insurance fraud detection model, the data to be detected is input into the medical insurance fraud detection model, and a fraud detection result output by the medical insurance fraud detection model is obtained.

[0034] In this way, by using real sample data to train the generative adversarial network, the generator in the generative adversarial network can generate synthetic sample data, and the synthetic sample data generated by the generator is screened through the particle swarm optimization algorithm to obtain screened synthetic sample data with better global diversity and local density. Together with the real sample data, it constitutes a training sample set to train the classification model. In this way, the minority classes in the real sample data can be expanded, and the category balance of the training sample set can be improved, thereby improving the performance of the trained medical insurance fraud detection model and achieving the effect of improving the accuracy of medical insurance fraud detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 It is a flow chart of the medical insurance fraud detection method based on particle swarm optimization generative adversarial network provided by the present invention.

[0037] Figure 2 This is a flowchart of the model training process in the medical insurance fraud detection method based on particle swarm optimization generative adversarial network provided by the present invention.

[0038] Figure 3 It is a structural diagram of a medical insurance fraud detection device based on a particle swarm optimization generative adversarial network provided by the present invention.

[0039] Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0040] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0041] It will be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0042] It should also be understood that the terms used in the present specification are only for the purpose of describing particular embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0043] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0044] As used in this specification and the appended claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0045] The following combination Figure 1-Figure 2 The present invention describes the medical insurance fraud detection method based on particle swarm optimization generative adversarial network. Figure 1 As shown in FIG, the medical insurance fraud detection method based on particle swarm optimization generative adversarial network includes the following steps:

[0046] S110: Inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, where the synthetic sample data includes synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected. The generative adversarial network also includes a discriminator. The training data of the generative adversarial network includes real sample data, where the real sample data includes real sample data to be detected and fraud labels corresponding to the real sample data to be detected.

[0047] S120. Screening the synthetic sample data based on a first particle swarm optimization algorithm to obtain a plurality of screened synthetic sample data. In the first particle swarm optimization algorithm, each particle corresponds to a sample set, and the sample set includes a plurality of synthetic sample data. The fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set.

[0048] S130, combining the plurality of screened synthetic sample data with the plurality of real sample data to obtain a training sample set, and training a classification model based on the training sample set to obtain a medical insurance fraud detection model;

[0049] S140: Input the data to be detected into the medical insurance fraud detection model to obtain the fraud detection result output by the medical insurance fraud detection model.

[0050] Health insurance fraud refers to the act of fabricating false medical data to claim from health insurance institutions (such as insurance companies selling commercial health insurance). Health insurance fraud detection aims to identify whether there is false data in health insurance claim data. In other words, health insurance fraud detection can be viewed as a binary classification problem, dividing the medical data used for claims into two categories: those with false data and those without. The present invention provides a health insurance fraud detection method based on a particle swarm optimization generative adversarial network. By training the generative adversarial network with real sample data, the generator in the generative adversarial network can generate synthetic sample data. The synthetic sample data generated by the generator is then filtered using a particle swarm optimization algorithm to obtain a training sample set with improved global diversity and local density. This is then used to train a classification model. This method can expand the minority classes in the real sample data, improve the class balance of the training sample set, and thus improve the performance of the trained health insurance fraud detection model, thereby achieving the effect of improving the accuracy of health insurance fraud detection.

[0051] Real sample data is constructed based on real medical insurance claim data, and the medical insurance fraud labels in the real sample data are obtained by manual annotation. Medical insurance claim data includes many different types of features, such as drug combinations, diagnostic codes, etc. In one possible implementation method, all types of features in the medical insurance claim data can be used for medical insurance fraud detection. However, this method will cause the data that the model needs to analyze to be high-dimensional and sparse, resulting in the model being susceptible to noise interference, low training efficiency and easy to overfit, and difficult to capture the implicit correlation patterns of fraudulent behavior. In another possible implementation method of the method provided by the present invention, the massive redundant features included in the medical insurance claim data are screened. Specifically, before the random noise is input into the generator of the trained generative adversarial network, it includes:

[0052] Obtain real medical insurance data, which includes various types of features;

[0053] Using the second particle swarm optimization algorithm, a target feature type is determined from multiple types of features, wherein a particle in the second particle swarm optimization algorithm corresponds to a feature type set, the feature type set includes at least one feature type, and the fitness value of the particle in the second particle swarm optimization algorithm reflects the discrimination ability and independence of the feature type in the feature type set corresponding to the particle;

[0054] Construct real sample data based on the target feature type.

[0055] By using the second particle swarm optimization algorithm to screen feature types and determine the target feature types to construct real sample data, the discriminability of features and redundancy are dynamically balanced, the robustness of feature selection under high-dimensional coefficient data is enhanced, and the problem of insufficient semantic expression ability in complex data is solved.

[0056] like Figure 2 As shown in the figure, when using the second particle swarm optimization algorithm to screen feature types, the particle swarm is first initialized: the particle swarm size is defined as N particles. The position of each particle represents a feature type set, which can be represented by a binary code. The number of digits in the code represents the total number of feature types in the medical insurance data. When the position corresponding to a feature type is 1, it indicates that the feature type is selected, and 0 indicates that the feature type is not selected. The particle position and velocity are randomly initialized, and the maximum number of iterations is set. In each iteration, the fitness value of each particle is calculated, and the particle is retained or updated based on the fitness value.

[0057] In the second particle swarm optimization algorithm, the fitness value of a particle reflects the discrimination ability and independence of the feature type in the feature type set corresponding to the particle. Specifically, the fitness value of a particle in the second particle swarm optimization algorithm is determined based on the first formula, which is:

[0058] ;

[0059] in, Indicates the fitness value of the particle whose corresponding feature type set is X, 、 、 is the weight coefficient, represents the mutual information between X and the fraud label set Y, , n is the total number of samples, Represents the value vector of the feature type of the i-th sample in the feature type set X, is the fraud label of the i-th sample, Represents the average value vector of the feature type of all samples in the feature type set X, represents the mean fraud label of all samples, represents the number of feature types in the feature type set X, and are the standard deviation vectors of X and Y respectively.

[0060] The samples in the above formula refer to real medical insurance data that has not undergone feature selection. Each real medical insurance data entry includes feature values ​​for various feature types and corresponding fraud labels. For each feature type set, because feature types are selected within the feature type set, the feature type value vectors for each sample in different feature type sets are different. For example, suppose there are four feature types (A, B, C, and D), and there are two samples: A and B. Feature type set C contains only feature types A, B, and C. In sample A, the values ​​of A, B, C, and D are ab, c, and d, respectively, while in sample B, the values ​​of A, B, C, and D are ef, g, respectively. Then, the feature type value vector for sample A in feature type set C1 is (a, b, c), and the feature type value vector for sample B in feature type set C1 is (e, f, g). The average feature type value vector for all samples in feature set C1 is ((a+e) / 2, (b+f) / 2, (c+g) / 2). Similarly, if the feature types included in the feature type set C2 are A and B, then the value vector of the feature type of sample A in the feature type set C2 is (ab), and so on.

[0061] In each iteration, based on the particle's fitness value, the particle's speed and position are updated. The particle speed update formula is:

[0062] ;

[0063] The particle position update formula is:

[0064] ;

[0065] In the particle velocity update formula, the superscript t represents the value in the tth iteration, and the position of each particle is It is a binary vector, 0 means that the feature type corresponding to the position is not selected, and 1 means that the feature type corresponding to the position is selected; w is the inertia weight, which is used to balance global search and local optimization. 、 are learning factors, which are used to control the degree to which particles follow the individual optimum and the group optimum respectively; 、 It is a random number in [0,1], which ensures the diversity of the feature type set through random perturbation; Represents the set of feature types with the highest fitness found by the particle during the iteration process (binary vector); It is the feature type set (binary vector) with the highest fitness found by the entire particle swarm during the iteration process; A probability value between 0 and 1 can be obtained to determine the position update of the particle.

[0066] After the iteration is completed, the feature type set corresponding to the global optimal particle is selected to construct the real sample data. The feature type set is iteratively screened by optimizing the multi-objective optimization strategy through the second particle algorithm, dynamically balancing the feature discrimination ability and redundancy suppression, and significantly enhancing the semantic expression ability of high-dimensional sparse data. Figure 2 As shown, after constructing multiple real sample data, the generative adversarial network is trained using these real sample data so that the generator in the generative adversarial network can generate synthetic sample data similar to the real sample data.

[0067] Generative Adversarial Network (WGAN) consists of a generator and a discriminator (also called Figure 2 The input of the generator is random noise, such as a 100-dimensional random noise vector z~N(0,1), and the output is the node feature. , node features The discriminator inputs the node features, which determine whether the data is generated by the generator or real. The discriminator outputs a real-number score, which reflects whether the data is real or generated. The network architecture for the discriminator and generator can adopt existing network structures, such as three fully connected layers.

[0068] The goal of training the generator is to generate synthetic sample data that is as close to the real sample data as possible, and the goal of training the discriminator is to maximize the difference in scores between the real sample data and the sample data generated by the generator. Based on these training goals, the loss functions of the generator and discriminator can be constructed.

[0069] Specifically, the loss function of the generator should be constructed based on minimizing the mean score of the discriminator on the synthetic samples. The loss function of the generator can be expressed as:

[0070] ;

[0071] in: Indicates that the noise vector z follows the distribution P z (z) (such as the expected value of the standard normal distribution), It represents the score given by the discriminator to the samples generated by the generator G based on the noise z.

[0072] The loss function of the discriminator should be constructed based on maximizing the difference between the real sample score and the synthetic sample score. The loss function of the discriminator can be expressed as:

[0073] ;

[0074] in, represents the score of the sample generated by the discriminator for the real sample x, It represents the expected value of the distribution that the real sample follows.

[0075] like Figure 2 As shown in Figure 2, the training iterative cycle of the generative adversarial network includes:

[0076] 1. Generate noise samples.

[0077] 2. Generate fake data.

[0078] 3. Calculate the discriminator loss .

[0079] 4. Use gradient descent method based on loss Update the parameters of the discriminator.

[0080] 5. Calculate the total loss of the generator .

[0081] 6. Use gradient descent to update the parameters of the generator based on the generator’s loss.

[0082] 7. Repeat the above steps until the preset number of iterations is reached or other stopping conditions are met.

[0083] After the training of the generative adversarial network is completed, the generator therein is used to generate synthetic sample data. In the method provided by the present invention, the generated synthetic sample data is also screened to obtain screened synthetic sample data with better global diversity and local density, so as to improve the data quality of the training data set of the medical insurance fraud detection model. When generating synthetic sample data, the fraud label corresponding to the synthetic sample data is the minority class in the real sample data. In the medical insurance claim data, the data labeled as fraud is the minority class. The number of synthetic sample data is N times the difference between the number of majority class and minority class samples in the real sample data, and N is greater than or equal to 2. In this way, the minority class samples can be fully expanded. Specifically, the generator can be used to generate N gen =(N maj -N min ) 2 synthetic sample data, N maj and N min Representing the number of majority and minority class samples in the real sample data, respectively. Using the first particle swarm optimization algorithm, the synthetic sample data is screened to obtain multiple screened synthetic sample data. In the first particle swarm optimization algorithm, each particle corresponds to a sample set, which includes multiple synthetic sample data output by the generator. The position of each particle can be represented by a binary vector whose number of bits matches the total number of synthetic sample data. Each bit is 0 or 1, indicating whether the synthetic sample data corresponding to that bit is selected. The specific process of the first particle swarm optimization algorithm is described below.

[0084] First, a particle swarm is initialized. Then, for each particle in each iteration, its corresponding fitness value is calculated. In the method provided by the present invention, the fitness value of a particle in the first particle swarm optimization algorithm reflects the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set.

[0085] Specifically, the local density calculation formula in the fitness value of particles in the first particle swarm optimization algorithm is:

[0086] ;

[0087] ;

[0088] in, is the local density of the synthetic sample data s, is the estimated probability density of the synthetic sample data s, k is the total number of synthetic sample data, Represents the synthetic sample data s and the synthetic sample data The Euclidean distance between is the standard deviation of the real sample data; d is the feature dimension in the synthetic sample data.

[0089] The global diversity calculation formula in the fitness value of particles in the first particle swarm optimization algorithm is:

[0090] ;

[0091] in, represents the global diversity of the sample set corresponding to the particle, Represents a sample set The 1-Wasserstein distance between the distribution of and the distribution of the true sample data.

[0092] In each iteration, the particle speed and position are updated according to the particle fitness value, and the sample set with the highest fitness is selected. S opt .

[0093] During the synthetic sample data screening stage, the particle swarm optimization algorithm is introduced to perform density perception and diversity evaluation on the synthetic sample pool. Through the weighted fitness function of local density (based on kernel density estimation) and global diversity (based on Wasserstein distance), the optimal sample subset is dynamically selected to cover multimodal distribution while avoiding redundancy.

[0094] The method provided by the present invention realizes the coordinated optimization of feature space and synthetic sample quality through the global search capability and dynamic feedback of the particle swarm optimization algorithm in both the feature type selection and synthetic sample screening stages, and effectively solves the problems of noise interference and distribution deviation in high-dimensional sparse medical insurance data.

[0095] After determining the screened synthetic data, the screened synthetic data is combined with the real sample data into a training sample set. The classification model is trained based on the training sample set to obtain a medical insurance fraud detection model.

[0096] After training the medical insurance fraud detection model, the data to be detected is input into the medical insurance fraud detection model to obtain the medical insurance fraud detection results output by the medical insurance fraud detection model. The feature type of the data to be detected used to input into the medical insurance fraud detection model should be consistent with the feature type of the data in the training sample set used to train the medical insurance fraud detection model.

[0097] The following describes the medical insurance fraud detection device based on particle swarm optimization generative adversarial network provided by the present invention. The medical insurance fraud detection device based on particle swarm optimization generative adversarial network described below and the medical insurance fraud detection method based on particle swarm optimization generative adversarial network described above can be referred to each other. Figure 3As shown, the medical insurance fraud detection device based on particle swarm optimization generative adversarial network provided by the present invention includes a sample synthesis module 310, a sample screening module 320, a training module 330 and a detection module 340. Among them:

[0098] A sample synthesis module 310 is configured to input random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, where the synthetic sample data includes synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected. The generative adversarial network also includes a discriminator, and the training data for the generative adversarial network includes real sample data, where the real sample data includes real sample data to be detected and fraud labels corresponding to the real sample data to be detected.

[0099] A sample screening module 320 is configured to screen the synthetic sample data based on a first particle swarm optimization algorithm to obtain a plurality of screened synthetic sample data. In the first particle swarm optimization algorithm, each particle corresponds to a sample set, and the sample set includes a plurality of synthetic sample data. The fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set.

[0100] A training module 330 is configured to combine the plurality of screened synthetic sample data with the plurality of real sample data to obtain a training sample set, and train a classification model based on the training sample set to obtain a medical insurance fraud detection model;

[0101] The detection module 340 is used to input the data to be detected into the medical insurance fraud detection model to obtain the fraud detection results output by the medical insurance fraud detection model.

[0102] Figure 4 An example of a physical structure diagram of an electronic device is shown below. Figure 4As shown, the electronic device may include: a processor 410, a communications interface 420, a memory 430, and a communication bus 440. The processor 410, the communications interface 420, and the memory 430 communicate with each other via the communication bus 440. The processor 410 may call logic instructions in the memory 430 to execute a medical insurance fraud detection method based on a particle swarm optimization generative adversarial network. The medical insurance fraud detection method based on a particle swarm optimization generative adversarial network includes: inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, the synthetic sample data including synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected, the generative adversarial network also including a discriminator, the training data of the generative adversarial network including real sample data, the real sample data including real sample data to be detected and fraud labels corresponding to the real sample data to be detected; based on a first particle swarm optimization algorithm, the synthetic sample data are screened to obtain multiple screened synthetic sample data, in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set includes multiple synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set; combining the multiple screened synthetic sample data and the multiple real sample data to obtain a training sample set, training a classification model based on the training sample set to obtain a medical insurance fraud detection model; inputting the data to be detected into the medical insurance fraud detection model to obtain a fraud detection result output by the medical insurance fraud detection model.

[0103] Furthermore, the logic instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0104] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the medical insurance fraud detection method based on particle swarm optimization generative adversarial network provided by the above methods. The medical insurance fraud detection method based on a particle swarm optimization generative adversarial network includes: inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, the synthetic sample data including synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected, the generative adversarial network also including a discriminator, the training data of the generative adversarial network including real sample data, the real sample data including real sample data to be detected and fraud labels corresponding to the real sample data to be detected; based on a first particle swarm optimization algorithm, the synthetic sample data are screened to obtain multiple screened synthetic sample data, in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set includes multiple synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set; combining the multiple screened synthetic sample data and the multiple real sample data to obtain a training sample set, training a classification model based on the training sample set to obtain a medical insurance fraud detection model; inputting the data to be detected into the medical insurance fraud detection model to obtain a fraud detection result output by the medical insurance fraud detection model.

[0105] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the medical insurance fraud detection method based on particle swarm optimization generative adversarial network provided by the above methods. The medical insurance fraud detection method based on a particle swarm optimization generative adversarial network includes: inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, the synthetic sample data including synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected, the generative adversarial network also including a discriminator, the training data of the generative adversarial network including real sample data, the real sample data including real sample data to be detected and fraud labels corresponding to the real sample data to be detected; based on a first particle swarm optimization algorithm, the synthetic sample data are screened to obtain multiple screened synthetic sample data, in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set includes multiple synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set; combining the multiple screened synthetic sample data and the multiple real sample data to obtain a training sample set, training a classification model based on the training sample set to obtain a medical insurance fraud detection model; inputting the data to be detected into the medical insurance fraud detection model to obtain a fraud detection result output by the medical insurance fraud detection model.

[0106] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0107] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A medical insurance fraud detection method based on particle swarm optimization generative adversarial network, characterized in that: include: Inputting random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, wherein the synthetic sample data includes synthetic sample data to be detected and a fraud label corresponding to the synthetic sample data to be detected, the generative adversarial network also includes a discriminator, the training data of the generative adversarial network includes real sample data, the real sample data includes real sample data to be detected and a fraud label corresponding to the real sample data to be detected, and the fraud label corresponding to the synthetic sample data is a minority class in the real sample data; Based on a first particle swarm optimization algorithm, the synthetic sample data is screened to obtain a plurality of screened synthetic sample data, wherein in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set includes a plurality of the synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set; combining the plurality of screened synthetic sample data with the plurality of real sample data to obtain a training sample set, and training a classification model based on the training sample set to obtain a medical insurance fraud detection model; Inputting the data to be detected into the medical insurance fraud detection model to obtain the fraud detection result output by the medical insurance fraud detection model; Before inputting random noise into the generator of the trained generative adversarial network, the method includes: Acquiring real medical insurance data, where the real medical insurance data includes multiple types of features; Determining a target feature type from the multiple types of features using a second particle swarm optimization algorithm, wherein a particle in the second particle swarm optimization algorithm corresponds to a feature type set, the feature type set includes at least one feature type, and a fitness value of the particle in the second particle swarm optimization algorithm reflects the discrimination ability and independence of the feature type in the feature type set corresponding to the particle; Constructing the real sample data based on the target feature type; The fitness value of the particle in the second particle swarm optimization algorithm is determined based on the first formula; The first formula is: ； in, Indicates the fitness value of the particle whose corresponding feature type set is X, 、 、 is the weight coefficient, represents the mutual information between X and the fraud label set Y, , , n is the total number of samples, Represents the value vector of the feature type of the i-th sample in the feature type set X, is the fraud label of the i-th sample, Represents the average value vector of the feature type of all samples in the feature type set X, represents the mean fraud label of all samples, represents the number of feature types in the feature type set X, and are the standard deviation vectors of X and Y respectively, and the value vectors of the feature types of each sample in different feature type sets are different.

2. The medical insurance fraud detection method based on particle swarm optimization generative adversarial network according to claim 1 is characterized in that: The local density calculation formula of the fitness value of particles in the first particle swarm optimization algorithm is: ； ； in, is the local density of the synthetic sample data s, is the estimated probability density of the synthetic sample data s, k is the total number of the synthetic sample data, Indicates that the synthetic sample data s and the synthetic sample data The Euclidean distance between is the standard deviation of the real sample data; d is the feature dimension in the synthetic sample data; is a normalizing constant.

3. The medical insurance fraud detection method based on particle swarm optimization generative adversarial network according to claim 1 is characterized in that: The global diversity calculation formula in the fitness value of particles in the first particle swarm optimization algorithm is: ； in, represents the global diversity of the sample set corresponding to the particle, Represents the sample set The 1-Wasserstein distance between the distribution of and the distribution of the real sample data.

4. The medical insurance fraud detection method based on particle swarm optimization generative adversarial network according to claim 1 is characterized in that: The number of the synthetic sample data is N times the difference between the number of majority class samples and the number of minority class samples in the real sample data, where N is greater than or equal to 2.

5. A medical insurance fraud detection device based on particle swarm optimization generative adversarial network, characterized in that: include: A sample synthesis module is configured to input random noise into a generator of a trained generative adversarial network to obtain synthetic sample data output by the generator, wherein the synthetic sample data includes synthetic sample data to be detected and fraud labels corresponding to the synthetic sample data to be detected, the generative adversarial network further includes a discriminator, the training data of the generative adversarial network includes real sample data, the real sample data includes real sample data to be detected and fraud labels corresponding to the real sample data to be detected, and the fraud labels corresponding to the synthetic sample data are the minority class in the real sample data; a sample screening module, configured to screen the synthetic sample data based on a first particle swarm optimization algorithm to obtain a plurality of screened synthetic sample data, wherein in the first particle swarm optimization algorithm, each particle corresponds to a sample set, the sample set including a plurality of synthetic sample data, and the fitness value of each particle is determined based on the global diversity of the sample set corresponding to the particle and the local density of each synthetic sample data in the sample set; a training module, configured to combine the plurality of screened synthetic sample data with the plurality of real sample data to obtain a training sample set, and train a classification model based on the training sample set to obtain a medical insurance fraud detection model; A detection module, configured to input the data to be detected into the medical insurance fraud detection model and obtain a fraud detection result output by the medical insurance fraud detection model; Before inputting random noise into the generator of the trained generative adversarial network, the method includes: Acquiring real medical insurance data, where the real medical insurance data includes multiple types of features; Determining a target feature type from the multiple types of features using a second particle swarm optimization algorithm, wherein a particle in the second particle swarm optimization algorithm corresponds to a feature type set, the feature type set includes at least one feature type, and a fitness value of the particle in the second particle swarm optimization algorithm reflects the discrimination ability and independence of the feature type in the feature type set corresponding to the particle; Constructing the real sample data based on the target feature type; The fitness value of the particle in the second particle swarm optimization algorithm is determined based on the first formula; The first formula is: ； in, Indicates the fitness value of the particle whose corresponding feature type set is X, 、 、 is the weight coefficient, represents the mutual information between X and the fraud label set Y, , , n is the total number of samples, Represents the value vector of the feature type of the i-th sample in the feature type set X, is the fraud label of the i-th sample, Represents the average value vector of the feature type of all samples in the feature type set X, represents the mean fraud label of all samples, represents the number of feature types in the feature type set X, and are the standard deviation vectors of X and Y respectively, and the value vectors of the feature types of each sample in different feature type sets are different.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the medical insurance fraud detection method based on particle swarm optimization generative adversarial network as described in any one of claims 1 to 4 is implemented.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the medical insurance fraud detection method based on particle swarm optimization generative adversarial network as described in any one of claims 1 to 4 is implemented.

8. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the medical insurance fraud detection method based on particle swarm optimization generative adversarial network as described in any one of claims 1 to 4 is implemented.

Images