Intelligent underwriting risk analysis method, device and equipment and medium

Abstract

The invention relates to the technical field of intelligent decision making, can be applied to a financial science and technology business system platform, and discloses an underwriting risk intelligent analysis method, device, equipment and medium, and the method comprises the steps: obtaining riding data and rider data of a target rider, and generating a corresponding data feature vector according to the riding data and the rider data; constructing a target risk classification model according to the data feature vector; performing risk classification on the target rider by the target risk classification model to obtain a corresponding risk identifier; and analyzing an underwriting risk control value of the target rider according to the risk identifier and a risk coefficient corresponding to the data feature vector. According to the invention, intelligent optimization of the underwriting period can be realized, and the underwriting risk control accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent decision-making technology, and in particular to an intelligent analysis method, device, equipment and medium for insurance risk. Background Art

[0002] In the field of financial technology, the application scenarios of intelligent premium pricing methods based on dynamic factor analysis are gradually increasing, but the existing technologies still have many shortcomings in the efficiency and accuracy of dynamic premium pricing.

[0003] For example, in the fintech sector, existing cycling insurance pricing models only use basic mileage to quantify risk. When a rider requests a re-underwriting request due to a decrease in the proportion of nighttime orders, the system struggles to correlate this data with the actual ride-hailing data for premium pricing. Consequently, risk factor updates lag behind actual scenario changes, leading to inaccurate premium pricing.

[0004] Existing insurance pricing technologies are primarily based on static risk assessment frameworks, for example, constructing risk profiles based on limited dimensions such as "age range" and "occupational category." While this approach can achieve basic pricing, it lacks the ability to extract dynamic, multi-dimensional features at the data fusion level. Furthermore, it struggles to integrate trajectory data for risk analysis at the risk warning level. Existing models are prone to risk omissions and premium bias. Therefore, intelligently optimizing the underwriting cycle and improving premium pricing accuracy are pressing challenges. Summary of the Invention

[0005] The present invention provides an underwriting risk intelligent analysis method, device, equipment and medium, the main purpose of which is to solve the problem of inaccurate underwriting risk control value.

[0006] In a first aspect, to achieve the above-mentioned objectives, the present invention provides an intelligent underwriting risk analysis method, comprising:

[0007] Acquire riding data and rider data of a target rider, and generate a corresponding data feature vector according to the riding data and the rider data;

[0008] Constructing a target risk classification model based on the data feature vector;

[0009] The target risk classification model is used to classify the target riders and obtain corresponding risk identifications;

[0010] The underwriting risk control value of the target rider is analyzed based on the risk identifier and the risk coefficient corresponding to the data feature vector.

[0011] In a second aspect, the present invention further provides an insurance risk intelligent analysis device, comprising:

[0012] a feature generation module, configured to obtain riding data and rider data of a target rider, and generate a corresponding data feature vector based on the riding data and the rider data;

[0013] A model building module, used to build a target risk classification model based on the data feature vector;

[0014] A risk classification module, configured to classify the target rider by using the target risk classification model to obtain a corresponding risk identifier;

[0015] The underwriting risk control module is used to analyze the underwriting risk control value of the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector.

[0016] In a third aspect, the present invention further provides an electronic device, comprising:

[0017] at least one processor; and,

[0018] a memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned method for intelligent analysis of underwriting risks.

[0020] In a fourth aspect, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the above-mentioned method for intelligent analysis of underwriting risks.

[0021] The present invention converts multi-source heterogeneous data into unified and computable numerical features, eliminates dimensional differences, and clarifies the relationship between high-risk behavior characteristics (such as speeding frequency) and low-risk characteristics (such as the use of protective gear); by performing feature selection on data feature vectors, each feature selection has a clear numerical basis, which improves data interpretability, and optimizes computational efficiency and accelerates model convergence by training the target risk classification model through the target loss function; the target risk classification model combines the target data feature vector to perform precise risk identification, timely reflect the risk identification results, and can support differentiated underwriting strategies, optimize resource allocation, and reduce underwriting risks; according to the risk identification and the risk coefficient corresponding to the data feature vector, the target rider is subjected to underwriting risk control, which can achieve fine adjustment of the risk control value and improve the accuracy of the risk control value. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only 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 labor.

[0023] Figure 1 A schematic diagram of an application environment for an intelligent analysis method for insurance risk according to an embodiment of the present invention;

[0024] Figure 2 A flowchart of an intelligent analysis method for insurance risks provided by one embodiment of the present invention;

[0025] Figure 3 A schematic diagram of a process for constructing a target risk classification model based on the data feature vector provided in one embodiment of the present invention;

[0026] Figure 4 A schematic diagram of a module of an intelligent analysis device for insurance risks provided by one embodiment of the present invention;

[0027] Figure 5 A schematic diagram of the structure of an electronic device for implementing an intelligent analysis method for insurance risks provided by one embodiment of the present invention;

[0028] Figure 6 Another structural schematic diagram of an electronic device for implementing an intelligent analysis method for insurance risks provided by one embodiment of the present invention.

[0029] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the technical solutions of the present disclosure, and to fully understand and implement how the present disclosure applies technical means to solve technical problems and achieve the corresponding technical effects, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. The embodiments of the present disclosure and the various features in the embodiments can be combined with each other without conflict, and the technical solutions formed are all within the scope of protection of the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present disclosure.

[0031] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, apparatus, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] An embodiment of the present application provides an intelligent analysis method for underwriting risk, and the execution subject of the intelligent analysis method for underwriting risk includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the device provided by the embodiment of the present application. In other words, the intelligent analysis method for underwriting risk can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0033] The present invention provides an intelligent analysis method for insurance risk, which can be applied in the following situations: Figure 1In the application environment. Among them, the client communicates with the server through the network. The server can obtain multi-source data through the client, and eliminate dimensional differences by converting multi-source heterogeneous data into unified and computable numerical features, and at the same time clarify the relationship between high-risk behavior features (such as speeding frequency) and low-risk features (such as the use of protective gear); by performing feature selection on the data feature vector, each feature selection has a clear numerical basis, which improves data interpretability, and optimizes the computational efficiency and accelerates model convergence by training the target risk classification model through the target loss function; the target risk classification model combines the target data feature vector to perform precise risk identification, and promptly reflects the risk identification results. At the same time, it can support differentiated underwriting strategies, optimize resource allocation, and reduce underwriting risks; according to the risk identification and the risk coefficient corresponding to the data feature vector, the target rider is subjected to underwriting risk control, which can realize fine-grained adjustment of the underwriting risk value and improve the accuracy of the underwriting risk value, and finally output the target underwriting risk value and feedback it to the client. Among them, the client can be, but is not limited to, various personal computers, laptops, smart phones, tablets, and portable wearable devices. The server can be implemented with an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific examples.

[0034] Reference Figure 2 FIG. 1 is a flow chart of an intelligent underwriting risk analysis method according to an embodiment of the present invention. In this embodiment, the intelligent underwriting risk analysis method includes:

[0035] S1. Obtain riding data and rider data of a target rider, and generate a corresponding data feature vector according to the riding data and the rider data.

[0036] In an embodiment of the present invention, the riding data includes geographic location information, climate condition information, etc. The geographic location information refers to the city center or suburban road section, including traffic flow, independent electric vehicle lanes, etc.; the climate condition information refers to the frequency of extreme weather, weather forecast for the next 30 days, etc.

[0037] The rider data described in the present invention includes basic data, driving experience, vehicle data, physiological data, etc. The basic data includes rider age, rider education, etc.; the driving experience refers to the age of driving an electric bicycle, whether an accident has occurred, whether there is a violation record, etc.; the vehicle data refers to the brand, price, frequency of repairs of the brand, accident frequency, etc. of the electric bicycle; the physiological data includes the target rider's blood oxygen, heart rate, riding posture, etc.

[0038] The data feature vector mentioned in the present invention refers to the data constructed by feature extraction of riding data and rider data. For example, by analyzing the riding routes of target riders, including the target riders' high-frequency riding routes and riding time periods, and using a geographic information system to analyze the areas where traffic accidents frequently occur on commonly used routes, high-risk sections are defined, that is, the data feature vector can refer to a route risk feature vector.

[0039] It is also possible to cooperate with the traffic management department to obtain the location, status and timing information of the traffic lights, and combine it with the target rider's passing time to determine whether there is any traffic violation data of running a red light, that is, the data feature vector can refer to the violation judgment feature vector.

[0040] In an embodiment of the present invention, the riding data and the rider data can be obtained through public channels such as social media and sports platforms. The target rider may publicly share riding records through platforms such as Strava and Garmin Connect, and the riding activity list and activity details can be extracted through an API interface (such as the Strava API) to obtain the riding data and rider data of the target rider.

[0041] The present invention can also obtain relevant data through a GPS tracking device, such as when a target rider uses a bike computer (such as Garmin Edge, Wahoo) or a smart bicycle (such as Specialized Turbo Creo) to synchronize data through a GPS tracking device, thereby obtaining the target rider's riding data and rider data.

[0042] For example, obtaining the target rider's riding data and rider data in a medical and health scenario can be used for health monitoring and sports injury prevention. Wearable devices such as smart bracelets / watches (such as Garmin, Apple Watch) can be used to monitor heart rate, blood oxygen saturation, sleep data, etc. in real time, and acceleration sensors and gyroscopes can be used to analyze riding posture (such as cadence, body tilt angle), thereby preventing muscle fatigue or joint injuries.

[0043] Medical collaborative research, such as cooperation with sports medicine laboratories, can also be used to obtain riders' in-depth physiological data such as muscle electrical signals (EMG) and lactate thresholds, and then combine them with hospital physical examination reports (such as bone density and joint wear) to assess sports injuries.

[0044] In the fintech business field, the commercial value of the target rider can be assessed by crawling the rider's training mileage, climbing height, and average speed from the official website of the event, or by analyzing the rider's fan growth, sponsorship brand cooperation, and other cooperation through social media, to assist in the athlete's potential investment decision-making and provide financial institutions with insurance product design suggestions (such as event interruption insurance for rider injuries).

[0045] In the embodiment of the present invention, generating a corresponding data feature vector based on the riding data and the rider data includes:

[0046] Performing missing value processing on the riding data and the rider data to obtain first riding data and first rider data;

[0047] performing outlier detection on the first riding data and the first rider data to obtain second riding data and second rider data;

[0048] Deduplication is performed on the second riding data and the second rider data to obtain target riding data and target rider data.

[0049] Performing data feature extraction on the target riding data and the target rider data to obtain riding features and rider features;

[0050] Normalizing the riding feature and the rider feature to obtain a normalized riding feature and a normalized rider feature;

[0051] The normalized riding feature and the normalized rider feature are combined to obtain a data feature vector.

[0052] In detail, the present invention can use the dropna() and fillna() functions of Pandas to process missing values ​​of the riding data and rider data to solve the data missing problem and ensure data integrity.

[0053] Among them, the missing value processing includes deleting missing values, filling missing values, etc. The deleting missing values ​​is applicable to fields with a low missing ratio or non-critical fields. The filling missing values ​​is applicable to numerical data, including mean / median filling, mode filling, and interpolation. The mean / median filling is applicable to numerical data (such as riding speed), the mode filling is applicable to categorical data (such as the vehicle brands commonly used by riders), and interpolation methods such as linear interpolation (inferring missing values ​​based on previous and subsequent data).

[0054] In detail, the present invention can calculate the standard deviation distance between the data point and the mean through statistical methods such as the Z-score method. Usually, |Z|>3 is considered an abnormality. The IQR (interquartile range) method can also be used to preset Q1-1.5×IQR to Q3+1.5×IQR as the normal range to identify data points that deviate from the normal distribution, thereby correcting the outliers and replacing them with the nearest non-outlier value or median.

[0055] In detail, data can be converted into hash values ​​through a hash algorithm, and repeated hash values ​​are regarded as duplicate data. The data can also be sorted and deduplicated. After the data is sorted, adjacent duplicate items are traversed and deleted to eliminate duplicate data and avoid deviations in subsequent data analysis.

[0056] The present invention can also perform a data consistency check on the target riding data and the target rider data to ensure that the formats of all data fields are consistent.

[0057] In detail, key information is extracted from the target riding data and the target rider data to reduce the dimension and improve the efficiency of subsequent model construction. The data features are normalized. Min-Max normalization can be used to scale the data to the [0,1] interval, or Z-score standardization can be used to convert the data into a distribution with a mean of 0 and a standard deviation of 1, thereby eliminating the differences in different feature dimensions and accelerating model convergence.

[0058] The present invention can combine the rider features and riding features through polynomial features, interactive features, and automatic feature crossing methods. The polynomial features, such as riding speed × riding distance, combine weather conditions and takeaway time periods to reflect the order timeliness requirements faced by riders. The interactive features, such as the ratio of heart rate change rate to altitude climbing rate, and the automatic feature crossing can use deep learning models (such as Factorization Machines) to automatically learn high-order combinations, capture interactive information between features, and improve the expression ability of subsequent models.

[0059] In an embodiment of the present invention, multi-source heterogeneous data is converted into unified and computable numerical features, dimensional differences are eliminated, and the weight relationship between high-risk behavior features (such as speeding frequency) and low-risk features (such as the use of protective gear) is clarified.

[0060] S2. Construct a target risk classification model based on the data feature vector.

[0061] In an embodiment of the present invention, feature selection is performed on the data feature vectors, and the data feature vectors that meet the selection conditions are used as training data sets. The risk classification model is trained in combination with a preset target loss function to obtain a target risk classification model.

[0062] like Figure 3 As shown, in an embodiment of the present invention, the target risk classification model is constructed according to the data feature vector, including:

[0063] Perform feature selection on the data feature vector according to a preset risk label vector to obtain a target feature vector;

[0064] Constructing a training data set for a risk classification model based on the target feature vector and the risk label vector;

[0065] The risk classification model is trained according to the training data set and a preset target loss function to obtain a target risk classification model.

[0066] The objective loss function is shown in the following formula:

[0067]

[0068] Wherein, F represents the value of the target loss function, w represents the normal vector of the preset decision hyperplane, b represents the preset bias term, C represents the preset penalty parameter, n represents the total number of the training data set, i represents the sequence number of the training data set, ε i represents the preset slack variable corresponding to the i-th training data set.

[0069] In detail, the correlation coefficient (such as Pearson coefficient, mutual information) between each data feature vector and a preset risk label vector can be calculated by the Pearson coefficient, and feature selection is performed based on the correlation coefficient and the predefined target feature score.

[0070] The present invention can divide the target feature vector into a training set, a validation set, a test set, etc. in a ratio of 7:2:1. Stratified sampling can be used to ensure that the proportions of each risk identification are consistent to solve the problem of sample imbalance (low-risk samples are usually far more than high-risk samples).

[0071] Among them, linear models such as logistic regression can be used to train the risk classification model through the target loss function to ensure the recall rate of high risk identification and improve the model convergence efficiency.

[0072] In the embodiment of the present invention, the step of performing feature selection on the data feature vector according to the preset risk label vector to obtain the target feature vector includes:

[0073] Calculating a correlation score between each of the data feature vectors and a preset risk label vector;

[0074] Key features are selected for the data feature vector according to the relevance score and a predefined target feature score to obtain a target feature vector.

[0075] In detail, statistical tests (such as chi-square test and ANOVA) can be used to select features that are significantly correlated with the risk label vector, and recursive feature elimination (RFE) and feature importance ranking (such as based on random forest) can be used to reduce the number of features and reduce the dimension.

[0076] For example, in the field of financial technology business, the data feature vector can be features such as user transaction frequency, number of cycling equipment changes, and proportion of abnormal transaction amounts at night. Lasso regression can be used for feature sparsification, and low-weight feature coefficients can be compressed to 0 through L1 regularization. Features with non-zero coefficients such as the number of cross-city abnormal logins, the proportion of high-risk merchant consumption, and the frequency of equipment changes are retained as target feature vectors.

[0077] Among them, the target feature vector can be aligned with the risk label vector to form structured training data. Three types of labels are defined according to the user's historical behavior: good credit (0), potential risk (1), and high-risk fraud (2). The sliding window method is used to convert the user's riding-transaction coupling feature sequence in the past 30 days into a static vector, and sensitive fields such as transaction amount are binned to meet the security requirements of financial data.

[0078] In the embodiment of the present invention, feature selection is performed on the data feature vector, and each feature selection has a clear numerical basis, thereby improving data interpretability. The target risk classification model is trained by the target loss function to optimize computational efficiency and accelerate model convergence.

[0079] S3. The target risk classification model is used to classify the target rider's risk and obtain a corresponding risk identifier.

[0080] In an embodiment of the present invention, a risk value of the target rider is calculated according to a target risk classification model, and the obtained target risk value is compared with a preset risk classification threshold to obtain a risk identifier corresponding to the target rider.

[0081] In the embodiment of the present invention, the target risk classification model is used to perform risk classification on the target rider to obtain a corresponding risk identifier, including:

[0082] Calculate the risk value of the target rider according to a preset risk calculation formula to obtain a target risk value;

[0083] The risk identifier corresponding to the target rider is determined based on the target risk value and a preset risk classification threshold.

[0084] The risk calculation formula is specifically shown as follows:

[0085]

[0086] Wherein, f(x) represents the target risk value, b represents the preset bias term, n represents the total number of the training data sets, i represents the sequence number of the training data sets, α i represents the Lagrange multiplier corresponding to the preset i-th training data set, K(x i,x) represents the preset kernel function, x i represents the i-th data feature vector, and x represents the preset support vector.

[0087] In an embodiment of the present invention, determining the risk identifier corresponding to the target rider according to the target risk value and a preset risk classification threshold includes:

[0088] If the target risk value is greater than or equal to the preset high-risk classification threshold, the target rider is determined to be a high-risk rider;

[0089] If the target risk value is less than the high-risk classification threshold and greater than the preset medium-risk classification threshold, the target rider is determined to be a medium-risk rider;

[0090] If the target risk value is less than or equal to the medium risk classification threshold, the target rider is determined to be a low-risk identified rider.

[0091] For example, in a healthcare scenario, a smart healthcare platform needs to stratify the cardiovascular disease risk of target riders to optimize the allocation of health intervention resources. The target risk value of the target rider's target feature vector is calculated based on the target risk classification model. The preset risk classification thresholds can be low risk (<0.5), medium risk (0.5-1.5), and high risk (>1.5). If the target risk value of the target rider is 1.67>1.5, it is determined to be a high risk indicator.

[0092] In the FinTech scenario, an insurance institution needs to conduct premium assessments on target riders, such as cycling insurance, and dynamically price them based on the riders' risk identification to balance risk coverage and premium competitiveness. The target feature vector is input into a pre-trained target risk classification model (such as a deep neural network), and the risk identification is divided into 5 levels: L1 (extremely low risk) to L5 (extremely high risk). The target risk classification model outputs the risk identification result for the target rider, and the cycling insurance is priced based on the risk identification result, which reduces the high compensation caused by cycling accidents and improves the profitability of insurance products.

[0093] In an embodiment of the present invention, the target risk classification model integrates the target data feature vector for precise risk identification. The model is regularly iterated and trained to promptly reflect emerging risk patterns. At the same time, it can support differentiated underwriting strategies, optimize resource allocation, and reduce underwriting risks.

[0094] S4. Analyze the underwriting risk control value of the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector.

[0095] In an embodiment of the present invention, underwriting risk control is performed on the target rider based on the risk identification result and the risk coefficient corresponding to each data feature vector. The target rider can choose a suitable underwriting plan according to his or her own risk situation to provide personalized customized underwriting.

[0096] In the embodiment of the present invention, the underwriting risk control value may be the premium corresponding to different insurances.

[0097] In an embodiment of the present invention, analyzing the underwriting risk control value of the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector includes:

[0098] Mapping the risk identifier to a preset underwriting risk control strategy to obtain a first basic underwriting risk control value corresponding to the risk identifier;

[0099] Calculate the corresponding risk coefficient based on the data feature vector and the preset feature weight;

[0100] Determining a second basic underwriting risk control value corresponding to the data feature vector according to the risk coefficient;

[0101] Underwriting risk control is performed according to the first basic underwriting risk control value and the second basic underwriting risk control value to obtain a target underwriting risk control value.

[0102] In detail, the second basic underwriting risk control value is specifically shown in the following formula:

[0103]

[0104] Wherein, P2 represents the second basic underwriting risk control value, N represents the total number of the data feature vectors, s represents the sequence number of the data feature vectors, and x s represents the sth data feature vector, γ s Represents the preset feature weight corresponding to the sth data feature vector.

[0105] In detail, the risk identifier is mapped to a preset underwriting risk control value strategy, and each risk identifier corresponds to a basic underwriting risk control value, such as the basic underwriting risk control value of a low-risk rider is 100, the basic underwriting risk control value of a medium-risk rider is 200, and the basic underwriting risk control value of a high-risk rider is 300.

[0106] Based on the basic underwriting risk control value, adjustments are made based on the risk coefficient corresponding to the target rider's data feature vector. For example, the final underwriting risk control value can be fine-tuned based on the riding duration, the specific risks of the road section, etc.

[0107] For example, assuming that the target rider's feature vector X is classified as a high-risk category after the target risk classification model, the corresponding basic underwriting risk control value is 200 yuan / month. The target rider's feature vector also includes high-frequency night riding x1 and riding in high-accident-prone areas x2. The underwriting risk control value will be further increased according to the target feature vector, such as adding 20 yuan for each target feature vector. The final monthly underwriting risk control value of the target rider is 200+20*x1+20*x2.

[0108] In the financial insurance scenario, cycling insurance risk control is carried out on the target riders, and risk control is carried out by combining behavioral data with behavioral risks. Through precise control of high-risk riders, the high compensation caused by cycling accidents is reduced, and the profitability of insurance products is improved. Insurance companies can also provide differentiated product services and enhance their market competitiveness.

[0109] In an embodiment of the present invention, underwriting risk control is performed on the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector, so that the underwriting risk control value can be finely adjusted and the accuracy of the underwriting risk control value can be improved.

[0110] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0111] like Figure 4 FIG. 1 is a functional module diagram of an intelligent analysis device for insurance risks provided by one embodiment of the present invention.

[0112] In the embodiment of the present disclosure, an intelligent analysis device for insurance risk is provided, and the intelligent analysis device for insurance risk corresponds to the intelligent analysis method for insurance risk in the above embodiment. Figure 5 As shown, the underwriting risk intelligent analysis device 100 can be installed in an electronic device. According to the functions to be implemented, the underwriting risk intelligent analysis device 100 includes a feature generation module 101, a model construction module 102, a risk classification module 103, and an underwriting risk control module 104. The functional modules are described in detail as follows:

[0113] A feature generation module 101 is configured to obtain riding data and rider data of a target rider, and generate a corresponding data feature vector based on the riding data and the rider data;

[0114] A model building module 102 is used to build a target risk classification model based on the data feature vector;

[0115] The risk classification module 103 is used to classify the target rider's risk using the target risk classification model to obtain a corresponding risk identifier;

[0116] The underwriting risk control module 104 is used to analyze the underwriting risk control value of the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector.

[0117] In one embodiment, when generating a corresponding data feature vector based on the riding data and the rider data, the feature generation module 101 is configured to:

[0118] Performing missing value processing on the riding data and the rider data to obtain first riding data and first rider data;

[0119] performing outlier detection on the first riding data and the first rider data to obtain second riding data and second rider data;

[0120] Deduplication is performed on the second riding data and the second rider data to obtain target riding data and target rider data.

[0121] Performing data feature extraction on the target riding data and the target rider data to obtain riding features and rider features;

[0122] Normalizing the riding feature and the rider feature to obtain a normalized riding feature and a normalized rider feature;

[0123] The normalized riding feature and the normalized rider feature are combined to obtain a data feature vector.

[0124] In one embodiment, when constructing a target risk classification model based on the data feature vector, the model construction module 102 is configured to:

[0125] Perform feature selection on the data feature vector according to a preset risk label vector to obtain a target feature vector;

[0126] Constructing a training data set for a risk classification model based on the target feature vector and the risk label vector;

[0127] The risk classification model is trained according to the training data set and a preset target loss function to obtain a target risk classification model.

[0128] In one embodiment, when performing feature selection on the data feature vector according to the preset risk label vector to obtain the target feature vector, the model building module 102 is configured to:

[0129] Calculating a correlation score between each of the data feature vectors and a preset risk label vector;

[0130] Key features are selected for the data feature vector according to the relevance score and a predefined target feature score to obtain a target feature vector.

[0131] In one embodiment, when the risk classification module 103 performs risk classification on the target rider using the target risk classification model and obtains a corresponding risk identifier, it is configured to:

[0132] Calculate the risk value of the target rider according to a preset risk calculation formula to obtain a target risk value;

[0133] The risk identifier corresponding to the target rider is determined based on the target risk value and a preset risk classification threshold.

[0134] In one embodiment, when determining the risk identifier corresponding to the target rider based on the target risk value and a preset risk classification threshold, the risk classification module 103 is configured to:

[0135] If the target risk value is greater than or equal to the preset high-risk classification threshold, the target rider is determined to be a high-risk rider;

[0136] If the target risk value is less than the high-risk classification threshold and greater than the preset medium-risk classification threshold, the target rider is determined to be a medium-risk rider;

[0137] If the target risk value is less than or equal to the medium risk classification threshold, the target rider is determined to be a low-risk identified rider.

[0138] In one embodiment, when analyzing the underwriting risk control value of the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector, the underwriting risk control module 104 is configured to:

[0139] Mapping the risk identifier to a preset underwriting risk control strategy to obtain a first basic underwriting risk control value corresponding to the risk identifier;

[0140] Calculate the corresponding risk coefficient based on the data feature vector and the preset feature weight;

[0141] Determining a second basic underwriting risk control value corresponding to the data feature vector according to the risk coefficient;

[0142] Underwriting risk control is performed according to the first basic underwriting risk control value and the second basic underwriting risk control value to obtain a target underwriting risk control value.

[0143] In the present invention, the specific definition of an intelligent underwriting risk analysis device can be found in the definition of an intelligent underwriting risk analysis method described above and will not be repeated here. Each module in the aforementioned intelligent underwriting risk analysis device can be implemented in whole or in part via software, hardware, or a combination thereof. Each of these modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a computer device memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0144] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it implements the functions or steps on the service side of an intelligent underwriting risk analysis method.

[0145] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, memory, a network interface, a display screen, and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When executed by the processor, the computer program implements the client-side functions or steps of a method for intelligent analysis of underwriting risks.

[0146] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are performed:

[0147] Acquire riding data and rider data of a target rider, and generate a corresponding data feature vector according to the riding data and the rider data;

[0148] Constructing a target risk classification model based on the data feature vector;

[0149] The target risk classification model is used to classify the target riders and obtain corresponding risk identifications;

[0150] The underwriting risk control value of the target rider is analyzed based on the risk identifier and the risk coefficient corresponding to the data feature vector.

[0151] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and apparatuses can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the module division is merely a logical function division, and actual implementation may employ other division methods.

[0152] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional modules.

[0153] Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the claims are intended to be embraced therein. Any reference to a figure in a claim should not be construed as limiting the claim to which it relates.

[0154] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0155] In some implementations of this embodiment, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps of the method described in the above embodiment are implemented.

[0156] The readable storage medium of the present invention stores a computer program, which, when executed by a processor of an electronic device, can implement:

[0157] Acquire riding data and rider data of a target rider, and generate a corresponding data feature vector according to the riding data and the rider data;

[0158] Constructing a target risk classification model based on the data feature vector;

[0159] The target risk classification model is used to classify the target riders and obtain corresponding risk identifications;

[0160] The underwriting risk control value of the target rider is analyzed based on the risk identifier and the risk coefficient corresponding to the data feature vector.

[0161] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can be found in the relevant descriptions of the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0162] The computer-readable storage medium may also store at least one computer-executable program / instruction, such as a computer-readable instruction. Computer-readable storage media include, but are not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Computer-readable storage media may include, for example, read-only memory (ROM), a hard disk, a flash memory, etc. For example, a non-transitory computer-readable storage medium may be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions stored on the computer-readable storage medium, the various methods described above may be performed.

[0163] In addition, the computer device may also include (but is not limited to) a data bus, an input / output (I / O) bus, a display, and input / output devices (eg, keyboard, mouse, speaker, etc.).

[0164] The processor can communicate with external devices via an I / O bus via a wired or wireless network.

[0165] In one embodiment, the at least one computer executable instruction may also be compiled into or constitute a software product / computer program product, wherein one or more computer executable instructions are executed by a processor to perform the various functions and / or method steps in the embodiments described in the present technology.

[0166] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0167] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0168] In the embodiments provided in the present disclosure, it should be understood that the disclosed devices and methods may also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram may represent a module, a program segment or a part of a code, and the above-mentioned module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, may be implemented with a dedicated hardware-based system that performs the specified function or action, or may be implemented with a combination of dedicated hardware and computer instructions.

[0169] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. 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. 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, and should all be included in the scope of protection of the present invention.

[0170] It should be noted that if software tools or components other than those of our company appear in the embodiments of this application, they are only used for illustration and do not represent actual use.

Claims

1. An intelligent analysis method for underwriting risk, characterized in that: The method comprises: Acquire riding data and rider data of a target rider, and generate a corresponding data feature vector according to the riding data and the rider data; Constructing a target risk classification model based on the data feature vector; The target risk classification model is used to classify the target riders and obtain corresponding risk identifications; The underwriting risk control value of the target rider is analyzed based on the risk identifier and the risk coefficient corresponding to the data feature vector.

2. The intelligent analysis method for underwriting risk according to claim 1, characterized in that: The constructing of a target risk classification model according to the data feature vector includes: Perform feature selection on the data feature vector according to a preset risk label vector to obtain a target feature vector; Constructing a training data set for a risk classification model based on the target feature vector and the risk label vector; The risk classification model is trained according to the training data set and a preset target loss function to obtain a target risk classification model.

3. The intelligent analysis method for underwriting risk according to claim 2, characterized in that: The step of performing feature selection on the data feature vector according to the preset risk label vector to obtain a target feature vector includes: Calculating a correlation score between each of the data feature vectors and a preset risk label vector; performing key feature selection on the data feature vector according to the relevance score and a predefined target feature score to obtain a preliminary feature vector; Performing dimensionality reduction processing on the preliminary feature vector to obtain a feature vector after dimensionality reduction; The feature stability test is performed on the feature vector after dimensionality reduction, and the feature vector with stability greater than the set threshold is selected as the target feature vector.

4. The intelligent analysis method for underwriting risk according to claim 1, characterized in that: The target risk classification model is used to classify the target rider by risk, and obtains a corresponding risk identifier, including: Calculate the risk value of the target rider according to a preset risk calculation formula to obtain a target risk value; The risk identifier corresponding to the target rider is determined based on the target risk value and a preset risk classification threshold.

5. The intelligent analysis method for underwriting risk according to claim 1, characterized in that: Analyzing the underwriting risk control value of the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector includes: Mapping the risk identifier to a preset underwriting risk control strategy to obtain a first basic underwriting risk control value corresponding to the risk identifier; Calculate the corresponding risk coefficient based on the data feature vector and the preset feature weight; Determining a second basic underwriting risk control value corresponding to the data feature vector according to the risk coefficient; Underwriting risk control is performed according to the first basic underwriting risk control value and the second basic underwriting risk control value to obtain a target underwriting risk control value.

6. The intelligent analysis method for underwriting risk according to claim 4, characterized in that: The step of determining the risk identifier corresponding to the target rider according to the target risk value and a preset risk classification threshold includes: If the target risk value is greater than or equal to the preset high-risk classification threshold, the target rider is determined to be a high-risk rider; If the target risk value is less than the high-risk classification threshold and greater than the preset medium-risk classification threshold, the target rider is determined to be a medium-risk rider; If the target risk value is less than or equal to the medium risk classification threshold, the target rider is determined to be a low-risk identified rider.

7. The intelligent analysis method for underwriting risk according to claim 1, characterized in that: Generating a corresponding data feature vector according to the riding data and the rider data includes: Performing missing value processing on the riding data and the rider data to obtain first riding data and first rider data; performing outlier detection on the first riding data and the first rider data to obtain second riding data and second rider data; Deduplication is performed on the second riding data and the second rider data to obtain target riding data and target rider data; Performing data feature extraction on the target riding data and the target rider data to obtain riding features and rider features; Normalizing the riding feature and the rider feature to obtain a normalized riding feature and a normalized rider feature; The normalized riding feature and the normalized rider feature are combined to obtain a data feature vector.

8. An intelligent analysis device for insurance risk, characterized in that: The device comprises: a feature generation module, configured to obtain riding data and rider data of a target rider, and generate a corresponding data feature vector based on the riding data and the rider data; A model building module, used to build a target risk classification model based on the data feature vector; A risk classification module, configured to classify the target rider by using the target risk classification model to obtain a corresponding risk identifier; The underwriting risk control module is used to analyze the underwriting risk control value of the target rider based on the risk identifier and the risk coefficient corresponding to the data feature vector.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute an intelligent analysis method for underwriting risk as described in 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, an intelligent analysis method for underwriting risk as described in any one of claims 1 to 7 is implemented.

Images