Financial product recommendation method

Abstract

The invention relates to the technical field of finance, and particularly discloses a financial product recommendation method, which comprises the following steps: collecting user information to generate a data sample, calculating an abnormal sample and removing the abnormal sample, predicting and filling missing data, and forming a user portrait; analyzing the user portrait by using a neural network algorithm, and extracting preference features and risk tolerance quantized values of the user; establishing a database; generating a personalized financial product recommendation list for the user by using a recommendation algorithm according to the key information in the user portrait and the product information in the financial product library; and collecting feedback of the user to the recommendation result, generating comprehensive feedback data, and continuously optimizing the recommendation algorithm of the financial product. According to the method, the deep learning neural network and the regularization technology are utilized, the user features are deeply mined, and the accuracy and reliability of feature extraction are improved. According to the method, multiple recommendation algorithms are fused, multiple factors are considered, and the recommendation list better meeting the personalized requirements of the user and the market dynamics is generated.

Description

Technical Field

[0001] The present invention relates to the field of financial technology, and in particular to a method for recommending financial products. Background Art

[0002] In today's financial market, there are many types of financial products, and users face many problems when choosing financial products that suit them. Traditional financial product recommendation methods have great limitations. On the one hand, the data processing method is not perfect. When collecting user information, there is a lack of effective processing methods for abnormal data and missing data, resulting in inaccurate user portraits built based on these data. For example, some abnormal historical financial behavior data may interfere with the judgment of the user's true investment preferences, and missing data will cause information gaps in the portrait and fail to fully reflect the user's characteristics.

[0003] On the other hand, the recommendation algorithm is not accurate and intelligent enough. The simple content filtering algorithm only makes recommendations based on the feature matching between the product and the user portrait, ignoring the similarity between users and group behavior; although the collaborative filtering algorithm takes into account the similarity between users, it is insufficient in processing the complex features of user portraits and cannot fully combine key factors such as the user's risk tolerance to make personalized recommendations. Moreover, in the process of recommendation, traditional methods rarely dynamically update the key information of financial products, such as the yield fluctuation range and risk level, and cannot adapt to the dynamic changes of the market.

[0004] In addition, traditional recommendation methods lack effective use of user feedback and do not optimize the recommendation algorithm based on user feedback on the recommendation results, resulting in a low degree of fit between the recommended financial products and the actual needs of users.

[0005] Therefore, a method for recommending financial products has become an urgent problem to be solved. Summary of the invention

[0006] The purpose of the present invention is to provide a method for recommending financial products, which solves the technical problems in the prior art such as imperfect data processing, inaccurate recommendation algorithms, untimely product information updates, and inability to effectively utilize user feedback to optimize algorithms. Through a complete data processing flow, personalized recommendations that integrate multiple algorithms, and algorithm optimization based on user feedback, user characteristics are comprehensively and accurately analyzed, and a personalized recommendation list is generated for the user in combination with financial product information. The recommendation algorithm is continuously optimized through user feedback, thereby improving the accuracy and effectiveness of recommendations and meeting users' diverse financial product needs.

[0007] To achieve the above object, the present invention provides a technical solution: a method for recommending financial products, comprising the following steps:

[0008] S1. Collect the user's basic information, historical financial behavior data, and risk tolerance questionnaire filled in by the user and generate data samples. Use the Isolation Forest algorithm to calculate abnormal samples and remove them. Use the linear regression model to predict and fill in missing data to form a user profile.

[0009] S2. Analyze user portraits using neural network algorithms to extract user preference characteristics and risk tolerance quantification values;

[0010] S3. Establish a database containing detailed information of various financial products, wherein the database contains the product's rate of return, risk level, and investment period;

[0011] S4. Generate a personalized list of financial product recommendations for users based on key information in the user portrait and product information in the financial product database by combining content-based filtering algorithm with collaborative filtering algorithm;

[0012] S5. Collect user feedback on recommendation results, generate comprehensive feedback data, and continuously optimize the recommendation algorithm for financial products based on the comprehensive feedback data.

[0013] Further, in step S1, the basic information includes age, gender, occupation, and income level; the historical financial behavior data includes historical investment product types, investment amounts, transaction times, and transaction frequencies;

[0014] The collected data is processed using a data cleaning algorithm. Assume that the collected historical financial behavior data constitutes a data set D. For each sample in the data set , , where N is the number of samples; the Isolation Forest algorithm is used to calculate the anomaly score ; The average path length correction factor ;

[0015] Where n=N, , is Euler's constant, For sample Average path length among multiple isolated trees;

[0016] Setting Thresholds ,like , then determine It is an abnormal sample and is removed;

[0017] For missing data, linear regression models are used to predict and fill in the missing data, ultimately forming complete and accurate user portrait data.

[0018] Further, feature extraction step: In step S2, the neural network model is used to conduct in-depth analysis of the user portrait; the feature vector of the neural network input layer is set to , No. Layer Neurons =1,2,⋯,L, L is the number of network layers; input ,in It is The weight matrix of the layer has the dimension , It is The bias vector of the layer, dimension is ; Activation function Using the ReLU function, that is ;

[0019] The output layer L outputs y= , the output layer has neurons, corresponding to the extracted preference features and risk tolerance quantification ; Update the weight matrix through back propagation algorithm and gradient descent method and the bias vector , to minimize the mean square error loss function ,in is the number of training samples, is the true value, is the predicted value.

[0020] Furthermore, during the training of the neural network model, a regularization method is used to regularize the weight matrix Constraints are imposed to prevent the model from overfitting.

[0021] Further, in step S3, the yield of the product includes the expected annualized yield and the fluctuation range of the actual yield; the risk level is divided into low, medium and high risk levels, represented by the values ​​1, 2 and 3 respectively; the investment period includes short-term, medium-term and long-term, the short-term is within one year, the medium-term is one to three years, and the long-term is more than three years;

[0022] The ARIMA (p, d, q) model is used to analyze the historical yield time series of financial products. Combined with market macroeconomic indicators, the actual yield fluctuation range is updated through a multivariate linear regression model.

[0023] Furthermore, for the update of risk level, in addition to considering the product’s own factors, the risk level is re-evaluated through a logistic regression model in combination with the market risk index.

[0024] Furthermore, in step S4, based on the content filtering algorithm, the user portrait feature vector is set By preference characteristics and risk tolerance quantification Composition, financial product feature vector , calculate the cosine similarity , get the score vector based on content filtering , where m is the number of financial products;

[0025] Based on the collaborative filtering algorithm, let the user-product matrix be R, where Represents the rating of user v on product w; uses the risk tolerance quantification value in the user portrait and preference characteristics Cluster users and calculate the cosine similarity between user v and user k of the same category ,in and are the average ratings of user v and user k respectively;

[0026] Predict user v’s rating for product w ,in is a set of users of the same category and similar to user v, and the score vector based on collaborative filtering is obtained ;

[0027] Set weight ( ),and Quantify value based on risk tolerance Dynamic adjustment, fused score vector ,in ;

[0028] according to Sort the financial products and select the top K products with relatively high scores to generate the recommendation list RL, where .

[0029] Furthermore, the DBSCAN clustering algorithm is used to cluster financial products. The DBSCAN algorithm calculates the distance between samples and divides the samples whose distance is within a certain threshold and the number of neighborhood samples is greater than the minimum number of samples into the same category, thereby ensuring that the products in the recommendation list come from different categories and increasing product diversity.

[0030] Furthermore, in step S5, the feedback includes the user's click behavior on the recommended product, whether to purchase the recommended product, and the evaluation of the recommended product, and a user feedback matrix F is constructed, where represents the comprehensive feedback score of user v on the recommended product w;

[0031] Collect user browsing behavior data on recommended pages, including dwell time , browse order ; Get the comprehensive feedback value through weighted calculation ,in is the weight, which is determined according to the actual business importance;

[0032] Use Q-Learning algorithm to optimize the recommendation algorithm. Suppose the state space S consists of user portraits and the current recommendation list RL, and the action space A is different recommendation algorithm parameters.

[0033] In Status Take action Then, transfer to the new state And get reward r, reward r is based on the comprehensive feedback value Calculate, that is ;

[0034] Q Function represents the expected long-term cumulative reward of taking action a in state s, and the update formula is:

[0035] ,in is the learning rate, is the discount factor;

[0036] By continuously iteratively updating the Q function, the action that maximizes the Q value is selected, that is, the optimal recommendation algorithm parameters;

[0037] In the A / B test, users are randomly divided into two groups, and different versions of the recommendation algorithm are used for recommendation. The reward r is calculated based on the feedback data of the two groups of users, and then the Q function is updated to determine the optimal algorithm version.

[0038] The advantages of the present invention over the prior art are that the present invention constructs an accurate and complete user portrait through multi-dimensional data collection, data cleaning and preprocessing algorithms, laying a solid foundation for personalized recommendations.

[0039] The present invention utilizes deep learning neural networks and regularization technology to deeply mine user features and improve the accuracy and reliability of feature extraction.

[0040] The present invention combines multiple algorithms to dynamically update the financial product database, ensuring the timeliness and accuracy of product information and reflecting dynamic changes in the market.

[0041] The present invention integrates multiple recommendation algorithms and considers multiple factors to generate a recommendation list that better meets the user's personalized needs and market dynamics.

[0042] The present invention continuously optimizes the recommendation algorithm through multiple optimization algorithms and user feedback, improves the accuracy and effectiveness of recommendations, and enhances user experience and market competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of a method for recommending financial products of the present invention.

[0044] Figure 2 It is a flow chart of the steps of constructing a user portrait of a method for recommending financial products of the present invention.

[0045] Figure 3 It is a flow chart of the feature extraction steps of a financial product recommendation method of the present invention.

[0046] Figure 4 It is a flow chart of the database establishment steps of a financial product recommendation method of the present invention.

[0047] Figure 5 It is a flow chart of the steps of generating a recommendation list of a method for recommending financial products of the present invention.

[0048] Figure 6 It is a flow chart of algorithm optimization steps of a financial product recommendation method of the present invention. DETAILED DESCRIPTION

[0049] Various exemplary embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention.

[0050] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0051] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0052] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0053] Combined with Figure 1-6 , a method for recommending a financial product of the present invention is further described in detail.

[0054] A method for recommending financial products, comprising the following steps:

[0055] Steps to construct a user portrait: Collect basic information about users, which should at least include age, gender, occupation, and income level; collect historical financial behavior data about users, which should include but not be limited to historical investment product types, investment amounts, transaction times, and transaction frequencies; and obtain risk tolerance questionnaires filled out by users, and combine the above information to form a user portrait.

[0056] The collected data is processed using a data cleaning algorithm. Assume that the collected historical financial behavior data constitutes a data set D. For each sample in the data set , (N is the number of samples), the IsolationForest algorithm is used to calculate the anomaly score , where the average path length correction factor is (n=N, ,in is Euler's constant), For sample The average path length among multiple isolated trees. Set threshold ,like , then determine The samples are considered abnormal and removed.

[0057] For missing data, assume that the missing investment amount data , using the linear regression model Perform prediction filling and estimate the regression coefficients by least squares method , get the predicted value Fill in missing values ​​and finally form complete and accurate user portrait data U.

[0058] Feature extraction step: Use deep learning models such as neural networks to conduct in-depth analysis of user profile U. Suppose the neural network input layer feature vector is , No. layer( , L is the number of network layers) neurons, input ,in It is The weight matrix of the layer has the dimension , It is The bias vector of the layer, dimension is , activation function Using the ReLU function, that is The output layer L outputs y = , assuming the output layer has neurons, corresponding to the extracted preference features and risk tolerance quantification . Update the weight matrix through back propagation algorithm and gradient descent method and the bias vector , to minimize the mean square error loss function ,in is the number of training samples, is the true value, is the predicted value.

[0059] Database establishment steps: Establish a database containing detailed information of various financial products, which records the yield of the products in detail, including the expected annualized yield , Fluctuation range of actual yield ; Risk level According to industry standards or custom standards, it is divided into low, medium and high risk levels, represented by values ​​1, 2 and 3 respectively; investment period , including short-term (less than one year), medium-term (one to three years), and long-term (more than three years), represented by the values ​​1, 2, and 3 respectively.

[0060] Use the ARIMA (p, d, q) model to analyze the historical yield time series of financial products , ( is the length of the time series) and the original time series is subjected to d-order difference to obtain a stationary series ,in is the first-order difference. The ARIMA model represents , where B is the lag operator, , is the autoregressive part, is the moving average part, is a white noise sequence with a mean of 0 and a variance of . Estimate model parameters by using methods such as maximum likelihood estimation and , predicting future returns , update the expected annualized rate of return At the same time, combined with the market macroeconomic indicators E (such as GDP growth rate, inflation rate, etc.), through the multivariate linear regression model Update the actual yield fluctuation range , the regression coefficients are estimated by the least squares method .

[0061] Recommendation list generation steps: Based on key information in the user profile, such as the quantitative value of risk tolerance , preference characteristics , and the product information in the financial product library, and use the recommendation algorithm to generate a personalized financial product recommendation list for users. The recommendation algorithm uses a combination of content filtering algorithm and collaborative filtering algorithm. Based on the content filtering algorithm, let the user portrait feature vector By preference characteristics and risk tolerance quantification Composition, financial product feature vector , calculate the cosine similarity , get the score vector based on content filtering , where m is the number of financial products.

[0062] Based on the collaborative filtering algorithm, let the user-product matrix be R, where Indicates the rating of user v on product w (or purchase behavior, which can be expressed as 0 / 1). Quantify the risk tolerance value in the user portrait and preference characteristics Cluster users and calculate the cosine similarity between user v and user k of the same category ,in and are the average ratings of user v and user k respectively. Predict the rating of user v for product w ,in is a set of users of the same category and similar to user v, and the score vector based on collaborative filtering is obtained . Set the weight ( ),and Quantify value based on risk tolerance Dynamic adjustment, when When it is high, increase appropriately ;when When it is low, reduce it appropriately The fused score vector ,in .according to Sort the financial products and select the top K products with higher scores to generate the recommendation list RL, where .

[0063] Algorithm optimization steps: Collect user feedback on recommendation results, including user click behavior on recommended products, purchase behavior, evaluation of recommended products, etc., and construct user feedback matrix F, where Represents the comprehensive feedback score of user v on the recommended product w. At the same time, the browsing behavior data of users on the recommended page, such as the stay time, is collected , browse order Etc., and the comprehensive feedback value is obtained through weighted calculation:

[0064] ;in is the weight, which is determined according to the actual business importance. Using the Q-Learning algorithm to optimize the recommendation algorithm, the state space S is composed of user portraits (including the quantitative value of risk tolerance) and preference characteristics ) and the current recommendation list RL, and the action space A is composed of different recommendation algorithm parameters (such as the weights of content-based filtering and collaborative filtering ). In the state Take action Then, transfer to the new state And get reward r, reward r is based on the comprehensive feedback value Calculations, e.g. Q function represents the expected long-term cumulative reward of taking action a in state s, and the update formula is ,in is the learning rate, is the discount factor. By continuously iteratively updating the Q function, the action that maximizes the Q value (i.e., the optimal recommendation algorithm parameters) is selected. In the A / B test, users are randomly divided into two groups, and different versions of the recommendation algorithm (corresponding to different actions a) are used for recommendation. The reward r is calculated based on the feedback data of the two groups of users, and then the Q function is updated to determine the optimal algorithm version.

[0065] In the user portrait construction step, data cleaning and preprocessing also include one-hot encoding of categorical data (such as gender, occupation, product type, etc.) to meet the input requirements of subsequent machine learning algorithms.

[0066] In the feature extraction step, during the training of the neural network model, a regularization method (such as L1 or L2 regularization) is used to regularize the weight matrix Constraints are made to prevent the model from overfitting, and the L2 regularization term is ,in is the regularization parameter.

[0067] In the database establishment step, in addition to considering the product's own factors, the risk level is updated by combining the market risk index M through a logistic regression model. Corresponding to low, medium and high risk levels respectively, X is a related variable including product characteristics and market risk index M) re-evaluate the risk level .

[0068] In the step of generating the recommendation list, the diversity algorithm is based on the clustering method and uses the DBSCAN clustering algorithm to cluster financial products. The DBSCAN algorithm calculates the distance between samples. (such as Euclidean distance , the distance is within a certain threshold The number of samples in the neighborhood is greater than the minimum number of samples. The samples are divided into the same category, thus ensuring that the products in the recommendation list come from different categories and increasing product diversity.

[0069] In the algorithm optimization step, during the A / B test, statistical hypothesis tests (such as t-tests) are used to determine whether the performance differences of different versions of the recommendation algorithm are significant. :There is no difference in performance between the two algorithms, alternative hypothesis :The performance of the two algorithms is different, by calculating the t value (in are the means of the two groups of user feedback indicators, , are the variances of the two groups of user feedback indicators, are the number of users in two groups respectively), and compared with the critical value. If ( is the significance level), the null hypothesis is rejected and it is believed that the performance of the two algorithms is significantly different, thus more accurately determining the optimal algorithm version.

[0070] The specific implementation process of a financial product recommendation method of the present invention is as follows:

[0071] 1. User portrait construction steps

[0072] (1) Data collection:

[0073] The basic information of a user was collected: age 35, male, occupation middle-level management in an enterprise, and annual salary of 500,000 yuan.

[0074] Historical financial behavior data: In the past year, he has invested in stock funds (investment amount of 100,000 yuan, transaction time is X month of XXX year, and transaction frequency is once a month), bonds (investment amount of 200,000 yuan, transaction time is X month of XXXX year, and transaction frequency is once a quarter).

[0075] Risk tolerance questionnaire: This user self-assessed that his / her risk tolerance is medium.

[0076] (2) Data cleaning and preprocessing:

[0077] Outlier detection: Assume that the number of samples of historical financial behavior data collected is N = 100 (including the data of this user and other users). For each sample , using the Isolation Forest algorithm to calculate the anomaly score ,in , n=N, , For sample The average path length among multiple isolated trees. Set threshold ,like , then determine The samples are considered abnormal and removed.

[0078] Missing value filling: The user's investment amount data is complete and there are no missing values.

[0079] Categorical data processing: Perform one-hot encoding on gender (male is coded as [1, 0], female is coded as [0, 1]), occupation (middle-level management is coded as [0, 0, 1, 0, …], there are 10 occupational categories in total), and product type (stock funds are coded as [1,0, 0], bonds are coded as [0, 1, 0], etc.), ultimately forming complete and accurate user portrait data U.

[0080] 2. Feature extraction steps

[0081] (1) Neural network model construction:

[0082] Neural network input layer feature vector , (including age, income level, gender after one-hot encoding, occupation, etc.).

[0083] The number of network layers is L=3, and the first layer has neurons, and the second layer has neurons, the output layer L has neurons, corresponding to the extracted preference features and risk tolerance quantification .

[0084] enter , the activation function uses the ReLU function, that is .

[0085] Output layer y= .

[0086] (2) Model training:

[0087] The weight matrix is ​​regularized using L2 Constrained, the L2 regularization term is ,set up .

[0088] Update the weight matrix through back-propagation algorithm and gradient descent method and the bias vector , to minimize the mean square error loss function , the number of training samples M = 1000. Finally, the user's preference characteristics and risk tolerance quantification values ​​are obtained. (The higher the quantified value, the greater the risk tolerance).

[0089] 3. Database establishment steps

[0090] (1) Financial product information entry:

[0091] There is a product called "Steady Growth Fund" in the database, with an expected annualized rate of return , the actual rate of return fluctuation range , risk level (Medium Risk), Investment Term (Medium term, 1-3 years).

[0092] (2) Update of yield and risk level:

[0093] Return forecast: Use the ARIMA (p, d, q) model to predict the fund's historical return time series (T=100 time points) are analyzed, and the original time series is differentiated by d=1 order to obtain a stationary series , The ARIMA model represents , the model parameters are estimated by maximum likelihood estimation and other methods and , predicting future returns , update the expected annualized rate of return after prediction It is 6.5%.

[0094] Update of real yield fluctuation range: Combined with market macroeconomic indicators E (such as GDP growth rate, inflation rate, etc.), through the multivariate linear regression model Update the actual yield fluctuation range , the regression coefficients are estimated by the least squares method , , after update .

[0095] Risk level update: Combined with the market risk index M, through the logistic regression model ( Corresponding to low, medium and high risk levels respectively, X is a related variable including product characteristics and market risk index M) re-evaluate the risk level , the risk level after evaluation is still 2.

[0096] 4. Recommendation list generation steps

[0097] (1) Based on content filtering algorithm:

[0098] User portrait feature vector By preference characteristics and risk tolerance quantification constitute, (Here it is simplified to a two-dimensional vector).

[0099] Financial product "Steady Growth Fund" feature vector , (corresponding to characteristics such as risk preference and expected return).

[0100] Calculate cosine similarity , get the score based on content filtering .

[0101] (2) Based on collaborative filtering algorithm:

[0102] The user-product matrix R uses the risk tolerance quantification value in the user profile and preference characteristics Cluster users. The cosine similarity between user v and user k of the same category , predict user v's rating for "steady growth fund" , get the score based on collaborative filtering .

[0103] (3) Fusion and sorting:

[0104] because Higher, appropriate increase , set the weight .

[0105] The fusion score .

[0106] The above calculations are performed on all financial products and they are sorted. The top K=10 products with higher scores are selected to generate a recommendation list RL. The “Steady Growth Fund” is included in the recommendation list due to its higher score.

[0107] (4) Increased diversity:

[0108] The DBSCAN clustering algorithm is used to cluster financial products, with Euclidean distance Calculate the distance between samples and set the threshold , minimum number of samples , the samples whose distance is within the threshold and whose number of neighborhood samples is greater than the minimum number of samples are classified into the same category, ensuring that the products in the recommendation list come from different categories and increasing product diversity.

[0109] 5. Algorithm optimization steps

[0110] (1) User feedback collection:

[0111] The user's click behavior score for the recommended product "Steady Growth Fund" is 3 (out of 5 points), the non-purchase behavior score is 0, and the evaluation score is 4. The user feedback matrix F is constructed. (comprehensive rating).

[0112] At the same time, we collect the browsing behavior data and stay time of users on the recommended pages. Seconds, browsing order (3rd product viewed in order).

[0113] Obtain comprehensive feedback value through weighted calculation ,in , , ,but .

[0114] (2) Q-Learning algorithm optimization:

[0115] The state space S consists of user profiles (including risk tolerance quantification values and preference characteristics ) and the current recommendation list RL, and the action space A is composed of different recommendation algorithm parameters (such as the weights of content-based filtering and collaborative filtering different values ​​of ).

[0116] In Status Take action Then, transfer to the new state And get reward r, here .

[0117] Q Function represents the expected long-term cumulative reward of taking action a in state s, and the update formula is , where the learning rate , discount factor .

[0118] (3) A / B testing:

[0119] The users are randomly divided into two groups, using different versions of the recommendation algorithm (corresponding to different actions a, assuming that one group , another group ) for recommendation.

[0120] Number of users in the first group , feedback index mean ,variance ; Number of users in the second group , feedback index mean ,variance .

[0121] Calculate the t value, .

[0122] Defining the significance level , critical value ,because , then reject the original hypothesis and believe that the performance of the two algorithms is significantly different, thus more accurately determining the optimal algorithm version.

[0123] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A method for recommending financial products, characterized in that: The following steps are included: S1. Collect the user's basic information, historical financial behavior data, and risk tolerance questionnaire filled out by the user and generate data samples. Use the Isolation Forest algorithm to calculate abnormal samples and remove them. Use the linear regression model to predict and fill in missing data to form a user profile. S2. Analyze user portraits using neural network algorithms to extract user preference characteristics and risk tolerance quantification values; S3. Establish a database containing detailed information of various financial products, wherein the database contains the product's rate of return, risk level, and investment period; S4. Generate a personalized list of financial product recommendations for users based on key information in the user portrait and product information in the financial product database by combining content-based filtering algorithm with collaborative filtering algorithm; S5. Collect user feedback on recommendation results, generate comprehensive feedback data, and continuously optimize the recommendation algorithm for financial products based on the comprehensive feedback data.

2. A method for recommending financial products according to claim 1, characterized in that: In step S1, the basic information includes age, gender, occupation, and income level; the historical financial behavior data includes historical investment product types, investment amounts, transaction times, and transaction frequencies; The collected data is processed using a data cleaning algorithm. Assume that the collected historical financial behavior data constitutes a data set D. For each sample in the data set , , where N is the number of samples; the IsolationForest algorithm is used to calculate the anomaly score ; The average path length correction factor ; where n=N, , is Euler's constant, For sample Average path length in multiple isolated trees; setting threshold ,like , then determine It is considered as an abnormal sample and removed; For missing data, linear regression models are used to predict and fill in the missing data, ultimately forming complete and accurate user portrait data.

3. A method for recommending financial products according to claim 2, characterized in that: Feature extraction Steps: In step S2, the neural network model is used to conduct in-depth analysis of the user portrait; the neural network input layer feature vector is set to , No. Layer Neurons =1,2,⋯,L, L is the number of network layers; input ,in It is The weight matrix of the layer has the dimension , It is The bias vector of the layer, dimension is Activation function Using the ReLU function, that is ; The output layer L outputs y= , the output layer has neurons, corresponding to the extracted preference features and risk tolerance quantification ; Update the weight matrix through back propagation algorithm and gradient descent method and the bias vector , to minimize the mean square error loss function ,in is the number of training samples, is the true value, is the predicted value.

4. A method for recommending financial products according to claim 3, characterized in that: In the training process of the neural network model, the weight matrix is ​​regularized by using a regularization method. Constraints are imposed to prevent the model from overfitting.

5. A method for recommending financial products according to claim 4, characterized in that: In step S3, the yield of the product includes the expected annualized yield and the fluctuation range of the actual yield; the risk level is divided into low, medium and high risk levels, represented by the values ​​1, 2 and 3 respectively; the investment period includes short-term, medium-term and long-term, the short-term is within one year, the medium-term is one to three years, and the long-term is more than three years; The ARIMA (p, d, q) model is used to analyze the historical yield time series of financial products. Combined with market macroeconomic indicators, the actual yield fluctuation range is updated through a multivariate linear regression model.

6. A method for recommending financial products according to claim 5, characterized in that: When updating the risk level, in addition to considering the product’s own factors, the risk level is re-evaluated through a logistic regression model in combination with the market risk index.

7. A method for recommending financial products according to claim 6, characterized in that: In step S4, based on the content filtering algorithm, the user portrait feature vector By preference characteristics and risk tolerance quantification Composition, financial product feature vector , calculate the cosine similarity , get the score vector based on content filtering , where m is the number of financial products; Based on the collaborative filtering algorithm, let the user-product matrix be R, where represents the rating of user v on product w; Utilize the risk tolerance quantification value in the user profile and preference characteristics Cluster users and calculate the cosine similarity between user v and user k of the same category ,in and are the average ratings of user v and user k respectively; Predict user v’s rating for product w ,in is a set of users of the same category and similar to user v, and the score vector based on collaborative filtering is obtained ; Set weight ( ),and Quantify value based on risk tolerance Dynamic adjustment, fused score vector ,in ; according to Sort the financial products and select the top K products with relatively high scores to generate the recommendation list RL, where .

8. A method for recommending financial products according to claim 7, characterized in that: The DBSCAN clustering algorithm is used to cluster financial products. By calculating the distance between samples, samples whose distance is within a certain threshold and whose number of neighborhood samples is greater than the minimum number of samples are classified into the same category, ensuring that the products in the recommendation list come from different categories.

9. A method for recommending financial products according to claim 8, characterized in that: In step S5, the feedback includes the user's click behavior on the recommended product, whether to purchase the recommended product, and the evaluation of the recommended product, and a user feedback matrix F is constructed, where represents the comprehensive feedback score of user v on the recommended product w; Collect user browsing behavior data on recommended pages, including dwell time , browse order ; Obtain comprehensive feedback value through weighted calculation ,in is the weight, which is determined according to the actual business importance; Use Q-Learning algorithm to optimize the recommendation algorithm. Suppose the state space S consists of user portraits and the current recommendation list RL, and the action space A is different recommendation algorithm parameters. In Status Take action Then, transfer to the new state And get reward r, reward r is based on the comprehensive feedback value Calculate, that is ; Q Function represents the expected long-term cumulative reward of taking action a in state s, and the update formula is: ,in is the learning rate, is the discount factor; By continuously iteratively updating the Q function, the action that maximizes the Q value is selected, that is, the optimal recommendation algorithm parameters; In the A / B test, users are randomly divided into two groups, and different versions of the recommendation algorithm are used for recommendation. The reward r is calculated based on the feedback data of the two groups of users, and then the Q function is updated to determine the optimal algorithm version.

Images