Claim settlement risk prediction method and device, electronic equipment and storage medium

Abstract

The embodiment of the application provides a claim settlement risk prediction method and device, electronic equipment and storage medium, belongs to the artificial intelligence technical field, and is applied to the financial technology field. The method comprises the following steps: obtaining policy claim settlement data of a target policy; performing feature extraction on the policy claim settlement data to obtain policy claim settlement features, wherein the policy claim settlement features comprise claim settlement feature components; performing quantum feature coding on the policy claim settlement features to obtain quantum amplitude coding and quantum angle coding of each claim settlement feature component; and performing claim settlement risk prediction on the target policy according to each claim settlement feature component, each quantum amplitude coding and each quantum angle coding, so that the efficiency of risk prediction can be improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and is applied to the field of financial technology, particularly to a method, device, electronic device, and storage medium for predicting claims risk. Background Technology

[0002] In the fintech sector, insurance institutions use policy claims data to predict claims risks in order to reduce payout losses. Taking reinsurance excess loss business as an example, reinsurance institutions use policy claims data provided by the original insurers to predict claims risks in order to reduce excess loss risk. However, existing risk prediction models lack the data processing capabilities to handle massive amounts of high-dimensional policy claims data, resulting in low efficiency in claims risk prediction. Summary of the Invention

[0003] The main objective of this application is to provide a method, apparatus, electronic device, and storage medium for predicting claims risks, with the aim of improving the efficiency of claims risk prediction.

[0004] To achieve the above objectives, a first aspect of this application proposes a claims risk prediction method, the method comprising: Obtain the policy claims data for the target policy; Feature extraction is performed on the policy claims data to obtain policy claims features; wherein, the policy claims features include claims feature components, and the number of claims feature components is [number missing]. , It is a natural number greater than or equal to 1; The policy claim features are quantum feature encoded to obtain the quantum amplitude code and quantum angle code for each claim feature component; wherein, the quantum amplitude code has a first quantum encoding bit depth, the first quantum encoding bit depth being [value missing]. The quantum angle code has a second quantum code bit depth, which is 1. Claim risk is predicted for the target policy based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes to obtain a claim risk category; wherein, the claim risk category is used to indicate whether the target policy has or does not have a claim risk.

[0005] In some embodiments, after predicting the claim risk category of the target policy based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes, the method further includes: Obtain the task processing priority for the target policy based on the stated claim risk category; The task object is determined based on the task processing priority. The target policy is assigned to the task object for processing.

[0006] In some embodiments, the step of performing quantum feature encoding on the policy claim features to obtain the quantum amplitude encoding and quantum angle encoding of each claim feature component includes: The computational base state of each of the claim feature components is determined based on the policy claim features; For each of the claim feature components, the claim feature component is amplitude encoded according to the computational ground state to obtain the quantum amplitude code; Determine the rotation matrix for each component of the claim feature based on the policy claim features; For each of the claim feature components, the claim feature component is angularly encoded according to the rotation matrix to obtain the quantum angle code.

[0007] In some embodiments, the rotation matrix includes a horizontal rotation submatrix, a vertical rotation submatrix, and a phase rotation submatrix. The step of angularly encoding the claim feature components based on the rotation matrix to obtain the quantum angle code includes: The claims feature components are horizontally encoded based on the horizontal rotation sub-matrix and the preset ground state to obtain the horizontal angle code. The claim feature components are vertically encoded based on the vertical rotation sub-matrix and the preset ground state to obtain the vertical angle code. The claims feature components are phase-encoded according to the phase rotat submatrix and the preset ground state to obtain quantum phase coding; The horizontal angle code, the vertical angle code, and the quantum phase code are integrated to obtain the quantum angle code.

[0008] In some embodiments, the step of predicting the claim risk of the target policy based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes to obtain a claim risk category includes: Quantum entanglement is performed on each of the claimed feature components, each of the quantum amplitude codes, and each of the quantum angle codes to obtain the target quantum code; The target quantum code is assessed for claims risk using a claims risk prediction model to obtain the claims risk category.

[0009] In some embodiments, the step of quantum entanglement based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes to obtain the target quantum code includes: Calculate the component difference value between every two of the claimed feature components; Each of the quantum amplitude codes is integrated to obtain a reference amplitude code; Each of the quantum angle codes is integrated to obtain a reference angle code; The target quantum code is obtained by performing quantum entanglement based on each component difference value, the reference amplitude code, the reference angle code, each quantum amplitude code, and each quantum angle code.

[0010] In some embodiments, the step of obtaining the target quantum code by performing quantum entanglement based on each of the component difference values, the reference amplitude code, the reference angle code, each of the quantum amplitude codes, and each of the quantum angle codes includes: Quantum entanglement is performed based on the reference amplitude code and the reference angle code to obtain the first quantum code; Quantum entanglement is performed based on each of the quantum amplitude codes and each of the quantum angle codes to obtain a second quantum code; Quantum entanglement is performed based on each component difference value and each quantum angle code to obtain a third quantum code; The first quantum code, the second quantum code, and the third quantum code are integrated to obtain the target quantum code.

[0011] To achieve the above objectives, a second aspect of this application provides a claims risk prediction device, the device comprising: The acquisition module is used to acquire the policy claims data of the target policy; The feature extraction module is used to extract features from the policy claim data to obtain policy claim features; wherein, the policy claim features include claim feature components, and the number of claim feature components is [number missing]. , It is a natural number greater than or equal to 1; A quantum coding module is used to perform quantum feature encoding on the policy claim features to obtain a quantum amplitude code and a quantum angle code for each claim feature component; wherein, the quantum amplitude code has a first quantum coding bit depth, the first quantum coding bit depth being [value missing]. The quantum angle code has a second quantum code bit depth, which is 1. The risk prediction module is used to predict the claim risk of the target policy based on each of the claim feature components, each of the quantum amplitude codes and each of the quantum angle codes, and obtain the claim risk category; wherein, the claim risk category is used to indicate whether the target policy has or does not have claim risk.

[0012] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0013] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0014] The claims risk prediction method, device, electronic device, and computer-readable storage medium proposed in this application acquire policy claims data of a target policy to perform risk prediction based on the policy claims data. Feature extraction is performed on the policy claims data to extract meaningful and representative features, and data dimensionality reduction is achieved to obtain policy claims features. These policy claims features include... To address the insufficient model processing power caused by high-dimensional features, the policy claim features are quantum feature encoded, resulting in a quantum amplitude code and a quantum angle code for each claim feature component. The quantum amplitude code has a first quantum encoding bit depth, which is [number of bits]. This allows for the simultaneous carrying of data using n qubits. A dimensional vector, equivalent to a single quantum parallel operation In the next iteration, exponential dimensionality compression and computational acceleration are achieved. The quantum angle encoding has a second quantum encoding bit depth of 1. By mapping claim feature components to quantum angle codes using one quantum bit, continuous features requiring precise interpretation are represented, thereby improving the interpretability of prediction results. Claim risk prediction for the target policy is performed based on each claim feature component, each quantum amplitude code, and each quantum angle code, resulting in a claim risk category. This exponential compression and acceleration significantly improve the efficiency of claim risk prediction. Attached Figure Description

[0015] Figure 1 This is a flowchart of the claims risk prediction method provided in the embodiments of this application; Figure 2 yes Figure 1 The flowchart of step S130 in the process; Figure 3 yes Figure 2 The flowchart of step S240 in the text; Figure 4 yes Figure 1 The flowchart of step S140 in the middle; Figure 5 yes Figure 4The flowchart of step S410 in the middle; Figure 6 yes Figure 5 The flowchart of step S540 in the text; Figure 7 This is another flowchart of the claims risk prediction method provided in the embodiments of this application; Figure 8 This is a schematic diagram of the claims risk prediction device provided in the embodiments of this application; Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0017] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0019] In the fintech sector, insurance institutions use policy claims data to predict claims risks in order to reduce payout losses. Taking reinsurance excess loss business as an example, reinsurance institutions use policy claims data provided by the original insurers to predict claims risks in order to reduce excess loss risk. However, existing risk prediction models lack the data processing capabilities to handle massive amounts of high-dimensional policy claims data, resulting in low efficiency in claims risk prediction.

[0020] Based on this, embodiments of this application provide a claims risk prediction method, a claims risk prediction device, an electronic device, and a computer-readable storage medium, aiming to improve the efficiency of claims risk prediction.

[0021] The claims risk prediction method, claims risk prediction device, electronic device, and computer-readable storage medium provided in this application are specifically described through the following embodiments. First, the claims risk prediction method in this application embodiment is described.

[0022] The claims risk prediction method provided in this application relates to the field of artificial intelligence technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the claims risk prediction method, but is not limited to the above forms.

[0023] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0024] Figure 1 This is an optional flowchart of the claims risk prediction method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S110 to S140.

[0025] Step S110: Obtain the policy claims data for the target policy; Step S120: Extract features from the policy claim data to obtain policy claim features; wherein, the policy claim features include claim feature components, and the number of claim feature components is... , It is a natural number greater than or equal to 1; Step S130: Perform quantum feature encoding on the policy claim features to obtain the quantum amplitude encoding and quantum angle encoding of each claim feature component; wherein, the quantum amplitude encoding has a first quantum encoding bit depth, the first quantum encoding bit depth being... The quantum angle encoding has a second quantum encoding bit, which is 1. Step S140: Based on each claim feature component, each quantum amplitude code, and each quantum angle code, predict the claim risk of the target policy to obtain the claim risk category; wherein, the claim risk category is used to indicate whether the target policy has or does not have claim risk.

[0026] Steps S110 to S140 as illustrated in the embodiments of this application are performed by... Encoding a dimensional vector into the amplitude and angle of an n-qubit quantum state enables exponential compression and speedup. When using the amplitude and angle of a quantum state for claims risk prediction, the efficiency of risk prediction can be greatly improved.

[0027] In step S110 of some embodiments, policy claim data of the target policy is obtained from insurance business systems, databases, and other channels. The target policy refers to a policy that has already received a claim; the policy claim data comprises all structured information generated throughout the entire process of the target policy from the reporting of an accident to the receipt of compensation, including the type of insurance, insured information, policy information, accident information, and compensation information. The type of insurance can be auto insurance, property insurance, or liability insurance; insured information includes age, region, credit rating, etc.; policy information includes the sum insured, retention amount, reinsurance ratio, etc.; accident information includes time, location, and accident type, etc.; and compensation information includes the compensation amount and payment time, etc.

[0028] Data cleaning and standardization are performed on policy claims data to ensure data quality. This includes detecting missing values ​​and filling them using K-nearest neighbor interpolation, as well as deduplicating the data. Policy claims data comprises categorical variables (such as accident type) and continuous variables. One-hot encoding is used for categorical variables, and z-value standardization is applied to continuous variables (such as claim amount) to ensure consistent data distribution.

[0029] In step S120 of some embodiments, policy claim data is input into a feature extraction network for feature extraction to obtain policy claim features. The feature extraction network is a deep learning structure with feature extraction capabilities, such as a transformer model or a convolutional neural network. Policy claim features are meaningful and representative features extracted from policy claim data, which can be used to quantify the risk attributes throughout the entire process of claim settlement. Policy claim features include claim feature components; to facilitate subsequent quantum encoding, the number of claim feature components is set to... , It is a natural number greater than or equal to 1.

[0030] Quantum feature encoding is applied to policy claim features to map high-dimensional data vectors to quantum states, essentially fitting these high-dimensional data vectors into quantum states within a Hilbert space. Specifically, amplitude encoding is performed on the policy claim features to obtain the quantum amplitude code for each claim feature component. The quantum amplitude code has a first quantum encoding bit depth of n bits, which represents the number of qubits used for amplitude encoding of the claim feature component. Amplitude encoding efficiently represents high-dimensional feature vectors and supports parallel computation, improving the efficiency of claim risk prediction. Angle encoding is then performed on the policy claim features to obtain the quantum angle code for each claim feature component. The quantum angle code has a second quantum encoding bit depth of 1 bit, which represents the number of qubits used for angle encoding of the claim feature component. Angle encoding maps field semantics to interpretable rotation angles, improving the interpretability of the features.

[0031] Please see Figure 2 In some embodiments, step S130 may include, but is not limited to, steps S210 to S240: Step S210: Determine the calculation base state of each claim feature component based on the policy claim features; Step S220: For each claim feature component, the amplitude of the claim feature component is encoded according to the calculated ground state to obtain the quantum amplitude code; Step S230: Determine the rotation matrix for each claim feature component based on the policy claim features; Step S240: For each claim feature component, perform angular encoding on the claim feature component according to the rotation matrix to obtain quantum angular encoding.

[0032] In step S210 of some embodiments, the number of claim feature components in the policy claim features is determined. Determine the number of qubits For each claim feature component, obtain its sequential identifier within the policy's claim features. Represent this sequential identifier using the number of qubits to obtain the computational ground state of the claim feature component. The computational ground state is an n-bit binary representation based on the sequential identifier.

[0033] The sequence identifier 'i' indicates that the claim feature component is the i-th component in the insurance claim features. If the number of qubits is n and the sequence identifier is i, then the computational ground state representation of the i-th claim feature component is |i>, which means the n-bit binary representation of i-1. For example, if the number of qubits is 2 and the sequence identifier is 4, then the computational ground state is |11>.

[0034] In step S220 of some embodiments, a normalization factor is determined based on the policy claim characteristics, and each claim characteristic component is normalized according to the normalization factor. The normalized claim characteristic component is multiplied by the corresponding computational ground state to obtain the quantum amplitude encoding of the claim characteristic component. The normalization factor is expressed as: Q , , Where Q is the normalization factor; The number of claim feature components; Let k be the k-th claim feature component.

[0035] The quantum amplitude encoding of the i-th claim feature component is represented as: , in, This represents the i-th claim feature component after normalization. This represents the computational ground state of the i-th claim feature component.

[0036] Amplitude encoding can be used to represent the signal, requiring only n qubits. The insurance claims feature greatly saves quantum resources.

[0037] In step S230 of some embodiments, for each claim feature component, a rotation matrix of the claim feature component is calculated based on the order identifier of the claim feature component in the policy claim features and the claim feature component itself. The rotation matrix includes a horizontal rotation submatrix, a vertical rotation submatrix, and a phase rotation submatrix. The horizontal rotation submatrix is ​​a matrix that rotates around the x-axis, the vertical rotation submatrix is ​​a matrix that rotates around the y-axis, and the phase rotation matrix is ​​a matrix that rotates around the z-axis.

[0038] The horizontal rotation submatrix is ​​denoted as RX, and RX can be defined as: , The vertical rotation submatrix is ​​denoted as RY, and RY can be defined as: , The phase rotation submatrix is ​​denoted as RZ, and RZ can be defined as: , in, This represents the i-th claim feature component.

[0039] In step S240 of some embodiments, for each claim feature component, the claim feature component is angularly encoded according to the horizontal rotation submatrix, the vertical rotation submatrix and the phase rotation submatrix respectively, and the claim feature component is mapped to a quantum rotation angle to obtain a quantum angle code.

[0040] By converting insurance claims data into a quantum computing-processable format through amplitude and angle encoding, a quantum-ready foundation is laid for subsequent claims risk prediction. This opens a channel between traditional data and quantum computing capabilities, enabling insurance claims characteristics to be instantly processed. Parallel computation in the Widhillbert space achieves exponential speedup.

[0041] Steps S210 to S240 above map high-dimensional claims features to the amplitude and angle of quantum states, enabling the calculation of claims using n qubits. A single dimensional vector participates in a physical evolution simultaneously, saving exponential memory usage, and leveraging the parallelism of quantum computing allows for a single quantum parallel operation. This improved computational efficiency, thereby increasing the efficiency of risk prediction.

[0042] Please see Figure 3 In some embodiments, step S240 may include, but is not limited to, steps S310 to S340: Step S310: Horizontally encode the claim feature components based on the horizontal rotation submatrix and the preset ground state to obtain the horizontal angle code; Step S320: Vertically encode the claim feature components based on the vertical rotation submatrix and the preset ground state to obtain the vertical angle code; Step S330: Perform phase encoding on the claim feature components based on the phase rotator matrix and the preset ground state to obtain quantum phase encoding; Step S340: Integrate the horizontal angle code, vertical angle code, and quantum phase code to obtain the quantum angle code.

[0043] In step S310 of some embodiments, the preset ground state is a binary representation based on one qubit, including |0> and |1>. Multiplying the horizontal rotatable submatrix by the preset ground state yields the horizontal angle code of the claim feature components. The horizontal angle code is represented as: .

[0044] In step S320 of some embodiments, the vertical rotation submatrix is ​​multiplied by a preset ground state to obtain the vertical angle encoding of the claim feature components. The vertical angle encoding is represented as: .

[0045] In step S330 of some embodiments, the phase rotator matrix is ​​multiplied by a preset ground state to obtain the quantum phase code of the claim feature component. The quantum phase code is expressed as: .

[0046] In step S340 of some embodiments, the horizontal angle code, the vertical angle code, and the quantum phase code are multiplied to obtain the quantum angle code.

[0047] Steps S310 to S340 above, through horizontal encoding, vertical encoding and phase encoding, can cover a single claim feature component onto a three-dimensional sphere, which not only enhances the kernel's expressive power but also preserves the interpretability of the angle encoding.

[0048] Please see Figure 4 In some embodiments, step S140 may include, but is not limited to, steps S410 to S420: Step S410: Perform quantum entanglement based on each claim feature component, each quantum amplitude code, and each quantum angle code to obtain the target quantum code; Step S420: The target quantum code is assessed for claims risk using a claims risk prediction model to obtain the claims risk category.

[0049] In step S410 of some embodiments, each claim feature component, each quantum amplitude code, and each quantum angle code are input into an entanglement layer for quantum entanglement, thereby converting the insurance claim feature from a quantum state to an entangled state to obtain the target quantum code. Claim risk is jointly determined by multiple factors such as the number of claims and credit rating, and these factors have high-order, nonlinear coupling relationships. In order to capture all potential risk combinations in a timely manner and improve the accuracy of claim risk prediction, quantum entanglement is used to form nonlocal correlations between different qubits, capturing the complex nonlinear relationships between claim feature components and the joint risk effects caused by accident location and climate conditions. The entanglement layer is an operation sequence composed of multiple quantum gates. Quantum gates are linear operators that act on quantum states to change the amplitude and phase of the quantum states.

[0050] In step S420 of some embodiments, the target quantum code is input into the claims risk prediction model for claims risk assessment to obtain a claims risk category. The claims risk category indicates whether the target policy has a claims risk or not. The claims risk prediction model is a trainable or non-trainable framework that takes quantum states as input and claims risk categories as output. The claims risk prediction model may employ quantum variational classifiers (parameterized quantum circuits), quantum convolutional neural networks, etc.

[0051] Steps S410 to S420 above capture the nonlinear relationship between various claim feature components through the entanglement property of quantum computing, so as to improve the accuracy of claim risk prediction.

[0052] Please see Figure 5 In some embodiments, step S410 may include, but is not limited to, steps S510 to S540: Step S510: Calculate the component difference value between every two claim feature components; Step S520: Integrate each quantum amplitude code to obtain the reference amplitude code; Step S530: Integrate each quantum angle code to obtain the reference angle code; Step S540: Perform quantum entanglement based on the difference value of each component, the reference amplitude code, the reference angle code, each quantum amplitude code, and each quantum angle code to obtain the target quantum code.

[0053] In step S510 of some embodiments, to enable the claims risk prediction model to perceive the differences between claims feature components, any two claims feature components are subtracted to obtain the component difference value. For example, the two claims feature components are respectively represented as... and The component difference value is then expressed as .

[0054] In step S520 of some embodiments, each quantum amplitude code is added together to obtain a reference amplitude code. The reference amplitude code is represented as: , in, Indicates the reference amplitude code; This indicates the number of claim feature components.

[0055] In step S530 of some embodiments, each quantum angle code is multiplied to calculate a tensor product, resulting in a reference angle code. The reference angle code is represented as: , in, Indicates the reference angle encoding; Multiplication symbol; Indicates the first A quantum angle encoding.

[0056] In step S540 of some embodiments, each component difference value, reference amplitude code, reference angle code, each quantum amplitude code and each quantum angle code are input into the entanglement layer to perform quantum entanglement to obtain the target quantum code.

[0057] Through the above steps S510 to S540, the nonlinear relationship between each claim feature component can be fully captured, thereby improving the accuracy of claim risk prediction.

[0058] Please see Figure 6 In some embodiments, step S540 may include, but is not limited to, steps S610 to S640: Step S610: Perform quantum entanglement based on the reference amplitude code and the reference angle code to obtain the first quantum code; Step S620: Perform quantum entanglement based on each quantum amplitude code and each quantum angle code to obtain the second quantum code; Step S630: Perform quantum entanglement based on the difference value of each component and the quantum angle code to obtain the third quantum code; Step S640: Integrate the first quantum code, the second quantum code, and the third quantum code to obtain the target quantum code.

[0059] In step S610 of some embodiments, the reference amplitude code and the reference angle code are multiplied to calculate the tensor product, and the tensor product is input into the entanglement layer to perform global quantum entanglement to obtain the first quantum code.

[0060] In step S620 of some embodiments, for each claim feature component, the quantum amplitude code and quantum angle code of the claim feature component are multiplied to calculate the tensor product, and then the tensor product is input into the entanglement layer for local quantum entanglement to obtain a sub-code. All sub-codes are added together to obtain the second quantum code.

[0061] In step S630 of some embodiments, each component difference value is subjected to controlled rotation to obtain a preset code for the component difference value. Each preset code is input into the entanglement layer for local quantum entanglement to obtain a first sub-code. Each quantum angle code is input into the entanglement layer for local quantum entanglement to obtain a second sub-code. The first sub-code and the second sub-code are multiplied to obtain a third quantum code.

[0062] In step S640 of some embodiments, the first quantum code, the second quantum code, and the third quantum code are merged to obtain the target quantum code.

[0063] Steps S610 to S640 above can fully capture the dependencies between the feature components through quantum entanglement, thereby enhancing the feature representation capability.

[0064] Please see Figure 7 In some embodiments, after step S140, the method may also include, but is not limited to, steps S710 to S730: Step S710: Obtain the task processing priority of the target policy based on the claim risk category; Step S720: Determine the task object based on the task processing priority; Step S730: Assign the target policy to the task object for processing.

[0065] In step S710 of some embodiments, if the claims risk category indicates that the target policy has a claims risk, then the task processing priority of the target policy is determined to be the first priority. If the claims risk category indicates that the target policy does not have a claims risk, then the task processing priority of the target policy is determined to be the second priority. The first priority is higher than the second priority. Tasks with higher priorities are processed first.

[0066] In step S720 of some embodiments, the task object is the object for subsequent processing of the target policy. If the task processing priority is first priority, the task object is determined to be the first object. If the task processing priority is second priority, the task object is determined to be the second object. Taking reinsurance excess of loss business as an example, the first object is the external excess of loss layer of the reinsurance institution, and the second object is the internal excess of loss layer of the reinsurance institution. Taking claims review as an example, the first object is the policy review seat, and the second object is the policy storage seat.

[0067] In step S730 of some embodiments, the target policy is assigned to a first object or a second object for subsequent processing. For example, the target policy is assigned to an external excess loss layer or an internal excess loss layer for excess loss processing.

[0068] Through steps S710 to S730, target policies with claims risks can be prioritized for handling in order to control claims risks.

[0069] Please see Figure 8 This application also provides a claims risk prediction device that can implement the above-mentioned claims risk prediction method. The claims risk prediction device includes: The acquisition module 810 is used to acquire the policy claims data of the target policy; The feature extraction module 820 is used to extract features from policy claim data to obtain policy claim features; these features include claim feature components, the number of which is... , It is a natural number greater than or equal to 1; The quantum coding module 830 is used to perform quantum feature encoding on the policy claim characteristics to obtain the quantum amplitude code and quantum angle code for each claim characteristic component; wherein, the quantum amplitude code has a first quantum coding bit depth, the first quantum coding bit depth being [value missing]. The quantum angle encoding has a second quantum encoding bit, which is 1. The risk prediction module 840 is used to predict the claim risk of the target policy based on each claim feature component, each quantum amplitude code and each quantum angle code, and obtain the claim risk category; wherein, the claim risk category is used to indicate whether the target policy has or does not have claim risk.

[0070] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned claims risk prediction method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0071] Please see Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 910 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 920 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 920 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 920 and called and executed by the processor 910 using the claims risk prediction method of the embodiments of this application. The input / output interface 930 is used to implement information input and output; The communication interface 940 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 950 transmits information between various components of the device (e.g., processor 910, memory 920, input / output interface 930, and communication interface 940); The processor 910, memory 920, input / output interface 930 and communication interface 940 are connected to each other within the device via bus 950.

[0072] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described claims risk prediction method.

[0073] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0074] The claims risk prediction method, claims risk prediction device, electronic device, and computer storage medium provided in this application embodiment, through... Encoding a dimensional vector into the amplitude and angle of an n-qubit quantum state enables exponential compression and speedup. When using the amplitude and angle of a quantum state for claims risk prediction, the efficiency of risk prediction can be greatly improved.

[0075] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0076] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0077] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0078] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0079] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0080] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0081] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0082] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0083] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0084] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0085] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for predicting claims risk, characterized in that, The method includes: Obtain the policy claims data for the target policy; Feature extraction is performed on the policy claims data to obtain policy claims features; wherein, the policy claims features include claims feature components, and the number of claims feature components is [number missing]. , It is a natural number greater than or equal to 1; The policy claim features are quantum feature encoded to obtain the quantum amplitude code and quantum angle code for each claim feature component; wherein, the quantum amplitude code has a first quantum encoding bit depth, the first quantum encoding bit depth being [value missing]. The quantum angle code has a second quantum code bit depth, which is 1. Claim risk is predicted for the target policy based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes to obtain a claim risk category; wherein, the claim risk category is used to indicate whether the target policy has or does not have a claim risk.

2. The method according to claim 1, characterized in that, After predicting the claim risk category of the target policy based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes, the method further includes: Obtain the task processing priority for the target policy based on the stated claim risk category; The task object is determined based on the task processing priority. The target policy is assigned to the task object for processing.

3. The method according to claim 1, characterized in that, The step of performing quantum feature encoding on the policy claim features to obtain the quantum amplitude encoding and quantum angle encoding of each claim feature component includes: The computational base state of each of the claim feature components is determined based on the policy claim features; For each of the claim feature components, the claim feature component is amplitude encoded according to the computational ground state to obtain the quantum amplitude code; Determine the rotation matrix for each component of the claim feature based on the policy claim features; For each of the claim feature components, the claim feature component is angularly encoded according to the rotation matrix to obtain the quantum angle code.

4. The method according to claim 3, characterized in that, The rotation matrix includes a horizontal rotation submatrix, a vertical rotation submatrix, and a phase rotation submatrix. The step of performing angular encoding on the claim feature components based on the rotation matrix to obtain the quantum angle encoding includes: The claims feature components are horizontally encoded based on the horizontal rotation sub-matrix and the preset ground state to obtain the horizontal angle code. The claim feature components are vertically encoded based on the vertical rotation sub-matrix and the preset ground state to obtain the vertical angle code. The claims feature components are phase-encoded according to the phase rotat submatrix and the preset ground state to obtain quantum phase coding; The horizontal angle code, the vertical angle code, and the quantum phase code are integrated to obtain the quantum angle code.

5. The method according to claim 1, characterized in that, The process of predicting the claim risk of the target policy based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes, to obtain claim risk categories, includes: Quantum entanglement is performed on each of the claimed feature components, each of the quantum amplitude codes, and each of the quantum angle codes to obtain the target quantum code; The target quantum code is assessed for claims risk using a claims risk prediction model to obtain the claims risk category.

6. The method according to claim 5, characterized in that, The step of performing quantum entanglement based on each of the claim feature components, each of the quantum amplitude codes, and each of the quantum angle codes to obtain the target quantum code includes: Calculate the component difference value between every two of the claimed feature components; Each of the quantum amplitude codes is integrated to obtain a reference amplitude code; Each of the quantum angle codes is integrated to obtain a reference angle code; The target quantum code is obtained by performing quantum entanglement based on each component difference value, the reference amplitude code, the reference angle code, each quantum amplitude code, and each quantum angle code.

7. The method according to claim 6, characterized in that, The step of performing quantum entanglement based on each component difference value, the reference amplitude code, the reference angle code, each quantum amplitude code, and each quantum angle code to obtain the target quantum code includes: Quantum entanglement is performed based on the reference amplitude code and the reference angle code to obtain the first quantum code; Quantum entanglement is performed based on each of the quantum amplitude codes and each of the quantum angle codes to obtain a second quantum code; Quantum entanglement is performed based on each component difference value and each quantum angle code to obtain a third quantum code; The first quantum code, the second quantum code, and the third quantum code are integrated to obtain the target quantum code.

8. A claims risk prediction device, characterized in that, The device includes: The acquisition module is used to acquire the policy claims data of the target policy; The feature extraction module is used to extract features from the policy claim data to obtain policy claim features; wherein, the policy claim features include claim feature components, and the number of claim feature components is [number missing]. , It is a natural number greater than or equal to 1; A quantum coding module is used to perform quantum feature encoding on the policy claim features to obtain a quantum amplitude code and a quantum angle code for each claim feature component; wherein, the quantum amplitude code has a first quantum coding bit depth, the first quantum coding bit depth being [value missing]. The quantum angle code has a second quantum code bit depth, which is 1. The risk prediction module is used to predict the claim risk of the target policy based on each of the claim feature components, each of the quantum amplitude codes and each of the quantum angle codes, and obtain the claim risk category; wherein, the claim risk category is used to indicate whether the target policy has or does not have claim risk.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 7.

Images