WO2020034593A1 - Method and apparatus for processing missing feature in crowd performance feature prediction - Google Patents

Abstract

Disclosed are a method and apparatus for processing a missing feature in a crowd performance prediction, wherein same relate to the technical field of information and mainly aim to avoid deviations in the filling-in of a missing feature and avoid deviations in the association between a prediction result of a trained prediction model and a corresponding feature, thereby being capable of improving the prediction precision of the trained prediction model. The method comprises: acquiring existing features in a crowd performance prediction training set, and a preset mixture Gaussian model corresponding to a missing feature, wherein the preset mixture Gaussian model is composed of multivariate Gaussian distributions corresponding to the missing feature; according to the existing features and an expected maximum algorithm for the preset mixture Gaussian model, estimating the missing feature; and filling in the missing feature in the crowd performance prediction training set. The present application is suitable for processing a missing feature in a crowd performance prediction.

Description

Method and device for processing missing features in crowd performance feature prediction Technical field

This application claims priority from the Chinese patent application filed on August 13, 2018 with the Chinese Patent Office, application number 2018109185213, and application name "Method and Device for Processing Missing Features in Crowd Performance Feature Prediction", the entire contents of which are incorporated by reference. Incorporated in the application.

Background technique

In recent years, many industries have begun to pay attention to crowd performance, especially to identify the performance level of the crowd. By identifying and rewarding the target crowd, it can greatly promote the overall efficiency of the company. At present, the performance level of the crowd can usually be predicted by training the performance prediction model of the crowd. When training the prediction model, the larger the amount of data in the training set, the higher the prediction accuracy of the prediction model obtained by training. However, in practical applications, there are fewer complete training sets that can be used to train a crowd performance level prediction model, and in most cases there will be missing features in the training set. Therefore, in order to improve the prediction accuracy of a prediction model, it is usually necessary to fill or complete missing features.

At present, traditional filling methods such as special value interpolation and hot card interpolation are usually used to fill the missing features in the crowd performance prediction training set. However, the filling effect of the above filling method is limited by the distribution of features, and the missing features are usually non-random missing features, and the absence of non-random missing features is related to the features themselves. For example, in the income survey, neither the high-income group nor the low-income group will fill in specific income, so the lack of income level is related to the income level itself. If the missing features are filled in the above manner, there will be deviations in the filling of missing features, which will result in lower prediction accuracy of the trained prediction model.

Summary of the Invention

The present application provides a method and a device for processing missing features in the prediction of crowd performance characteristics, mainly to avoid deviations in the filling of missing features, and to avoid deviations in the association between the prediction results of the trained prediction model and the corresponding features. The prediction accuracy of the trained prediction model.

According to a first aspect of the present application, a method for processing missing features in a crowd performance prediction is provided, including:

Obtaining a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set, the preset mixed Gaussian model consisting of a multivariate Gaussian distribution corresponding to the missing features;

Estimating a value corresponding to the missing feature according to a maximum expectation algorithm of the existing feature and the preset mixed Gaussian model;

The values corresponding to the missing features are filled into the crowd performance prediction training set.

According to a second aspect of the present application, a missing feature processing device for crowd performance prediction is provided, including:

An obtaining unit, configured to obtain an existing feature and a preset mixed Gaussian model corresponding to the missing feature in the training set for predicting the performance of the crowd; the preset mixed Gaussian model is composed of a multivariate Gaussian distribution corresponding to the missing feature;

An estimation unit, configured to estimate a value corresponding to the missing feature according to a maximum expectation algorithm of the existing feature and the preset mixed Gaussian model;

The filling unit is configured to fill a value corresponding to the missing feature into the crowd performance prediction training set.

According to a third aspect of the present application, a computer non-volatile readable storage medium is provided, on which computer-readable instructions are stored, and when the computer-readable instructions are executed by a processor, the following steps are implemented:

Obtaining a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set, the preset mixed Gaussian model consisting of a multivariate Gaussian distribution corresponding to the missing features;

Estimating a value corresponding to the missing feature according to a maximum expectation algorithm of the existing feature and the preset mixed Gaussian model;

The values corresponding to the missing features are filled into the crowd performance prediction training set.

According to a fourth aspect of the present application, a computer device is provided, including a memory, a processor, and computer-readable instructions stored on the memory and executable on the processor. When the processor executes the computer-readable instructions, Implement the following steps:

Obtaining a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set, the preset mixed Gaussian model consisting of a multivariate Gaussian distribution corresponding to the missing features;

Estimating a value corresponding to the missing feature according to a maximum expectation algorithm of the existing feature and the preset mixed Gaussian model;

The values corresponding to the missing features are filled into the crowd performance prediction training set.

The method and device for processing missing features in the prediction of crowd performance characteristics provided by the present application are compared with the conventional filling methods, such as special value interpolation and hot card interpolation, which are currently used to fill the missing characteristics in the crowd performance prediction training set. The present application can obtain the existing features and the preset mixed Gaussian models corresponding to the missing features in the crowd performance prediction training set, and the preset mixed Gaussian models are composed of a multivariate Gaussian distribution corresponding to the missing features; The maximum expectation algorithm of the feature and the preset mixed Gaussian model estimates a value corresponding to the missing feature. At the same time, the values corresponding to the estimated missing features can be filled into the crowd performance prediction training set, so that the missing data can be filled based on the preset mixed Gaussian model corresponding to the missing features. Since the preset mixed Gaussian model is It consists of a multivariate Gaussian distribution corresponding to the missing feature, so it can ensure that the missing feature after filling reflects its distribution state and is related to itself, can avoid the bias of missing feature filling, and can avoid the prediction result and corresponding of the trained prediction model. There is a bias in the association between features, which can improve the prediction accuracy of the trained prediction model.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings described here are used to provide a further understanding of the present application and constitute a part of the present application. The schematic embodiments of the present application and the description thereof are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings:

FIG. 1 shows a flowchart of a method for processing missing features in a crowd performance prediction according to an embodiment of the present application;

FIG. 2 shows a flowchart of another method for processing missing features in a crowd performance prediction provided by an embodiment of the present application;

3 is a schematic structural diagram of a device for processing missing features in a crowd performance prediction provided by an embodiment of the present application;

FIG. 4 is a schematic structural diagram of another missing feature processing device in crowd performance prediction according to an embodiment of the present application; FIG.

FIG. 5 is a schematic structural diagram of a computer device according to an embodiment of the present application.

detailed description

Hereinafter, the present application will be described in detail with reference to the drawings and embodiments. It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

As the background technology, currently, traditional filling methods, such as special value interpolation and hot card interpolation, are usually used to fill the missing features in the crowd performance prediction training set. However, the filling effect of the above filling method is limited by the distribution of features, and the missing features are usually non-random missing features, and the absence of non-random missing features is related to the features themselves. For example, in the income survey, neither the high-income group nor the low-income group will fill in specific income, so the lack of income level is related to the income level itself. If the missing features are filled in the above manner, there will be deviations in the filling of missing features, which will result in lower prediction accuracy of the trained prediction model.

In order to solve the above problem, an embodiment of the present application provides a method for processing missing features in a crowd performance prediction. As shown in FIG. 1, the method includes:

101. Obtain a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set.

The preset mixed Gaussian model may be composed of a multivariate Gaussian distribution corresponding to the missing feature. The training set may include a crowd performance feature and a crowd performance level, and the missing feature in the embodiment of the present application may be a feature in the crowd performance feature. The performance characteristics of the crowd may include, but are not limited to, the average number of courses per month, the latitude and longitude of the work address, the number of Internet transaction products in a single month, and the level of Internet transactions in six months. For example, if there are 100 training samples, and 40 of the training samples do not have Internet transaction levels within six months, you can confirm that the "Internet transaction levels within six months" of the 40 training samples are missing features. Existing features can be "the average number of learning courses per month,

Work address latitude and longitude, number of Internet trading products in a single month, Internet transaction level within half a year "and" the average number of monthly courses, work address latitude and longitude, and the number of Internet trading products in a single month "among 40 training samples.

It should be noted that the probability distribution of the preset mixed Gaussian model can be expressed as follows:

If the missing feature in the embodiment of the present application is “Internet transaction level within half a year”, the multiple Gaussian distribution may be multiple categories of Internet transaction levels within half a year. According to the above formula, the category of Internet transaction levels within half a year may specifically be m.

102. Estimate a value corresponding to the missing feature according to a maximum expectation algorithm of the existing feature and the preset mixed Gaussian model.

The maximum expectation algorithm may be an iterative algorithm, and may include a maximum likelihood estimation calculation and an expectation calculation, and the maximum likelihood estimation calculation and the expectation calculation may be performed iteratively. In the embodiment of the present application, it is possible to iteratively calculate the distribution parameters of each meta-Gaussian distribution of the preset mixed Gaussian model by using the existing features and the maximum likelihood estimation calculation, and then calculate the missing by the expected calculation and the distribution parameters calculated by iteration. The probability that a feature belongs to each meta-Gaussian distribution, so that the Gaussian distribution to which the missing feature belongs can be estimated, and the value corresponding to the missing feature can be determined.

For example, if 60 training samples in 100 training samples have complete features and 40 training samples have missing features, you can estimate 40 training samples based on the existing features in the 60 training samples and a preset mixed Gaussian model. After the missing feature is estimated, and the value corresponding to the missing feature is estimated, by filling the value corresponding to the missing feature to the training set of performance prediction of the crowd, 100 training samples with complete features can be obtained.

103. Fill a value corresponding to the missing feature into the crowd performance prediction training set.

It should be noted that the value corresponding to the missing feature may be filled into the corresponding position of the crowd performance prediction training set according to the position of the existing feature. For example, training sample 3 is (the average number of courses per month is 80, the latitude and longitude of the work address (123.435, 41.819), and the Internet trading product in a single month is 100,). The missing feature in training sample 3 is "Internet trading level within half a year". In the training sample 3, when the corresponding value of "Internet Trading Level in Half a Year" is Internet Trading Level 2 in half a year, you can use the "average number of courses per month 80, work address latitude and longitude (123.435, 41.819), Internet trading products in a single month 100 ", filling the" Internet trading level 2 in half a year "into training sample 3, and obtaining (average number of courses per month 80, work address latitude and longitude (123.435, 41.819), Internet trading products 100 in a single month, Internet in half a year Transaction level 2).

The method for processing missing features in the prediction of crowd performance characteristics provided by the embodiments of the present application is compared with the conventional filling methods, such as special value interpolation, hot card interpolation, etc., which are used to fill the missing features in the crowd performance prediction training set. In the embodiment of the present application, a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set can be obtained, and the preset mixed Gaussian model is composed of a multivariate Gaussian distribution corresponding to the missing features; An existing feature and a maximum expected algorithm of the preset mixed Gaussian model estimate the missing feature. At the same time, the estimated missing features can be filled into the crowd performance prediction training set, so that the missing data can be filled based on a preset mixed Gaussian model corresponding to the missing features. The composition of the multivariate Gaussian distribution corresponding to the missing features can ensure that the missing features after filling reflect their distribution status and are related to themselves. It can avoid the bias of missing feature filling and can avoid the prediction results of the trained prediction model and the corresponding features. There is a bias in the association, which can improve the prediction accuracy of the trained prediction model.

Further, in order to better explain the process of missing features in the above-mentioned crowd performance prediction, as a refinement and extension of the foregoing embodiment, this embodiment of the present application provides another method for processing missing features in the crowd performance prediction, such as As shown in FIG. 2, the method includes:

201. Obtain a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set.

The preset mixed Gaussian model may be composed of a multivariate Gaussian distribution corresponding to the missing feature. In the embodiment of the present application, the existing features may exist in the form of feature vectors. If the features of the three dimensions of “average number of courses per month, latitude and longitude of work address, and number of Internet trading products in a single month” are uniquely marked for one and a half years Intra-Internet transaction level categories, the "average number of courses per month, latitude and longitude of work address, and number of Internet trading products in a single month" can be expressed in the form of feature vectors.

For the embodiment of the present application, in order to obtain the preset mixed Gaussian model, the method may further include: determining a multivariate Gaussian distribution corresponding to the missing feature; and constructing the preset mixed Gaussian model according to the multivariate Gaussian distribution.

For example, the missing feature is "Internet transaction level within half a year". Multiple categories corresponding to "Internet transaction level within half a year" are determined, and each category can correspond to a one-dimensional Gaussian distribution. The "average number of monthly courses, work address latitude and longitude" "The number of Internet trading products in a single month" as the observation sample vector, that is, the observed sample vector can be used to observe the multi-dimensional Gaussian distribution of missing features, and the Internet transaction level within six months can be divided into multiple categories, and then The weight, the corresponding mean vector, and the covariance matrix are used to construct a preset mixed Gaussian model corresponding to the Internet transaction level within the half year.

202. Estimate initial distribution parameters of each elementary Gaussian distribution of the preset hybrid Gaussian model according to the maximum likelihood estimation calculation of the existing feature and the maximum expected algorithm of the preset hybrid Gaussian model.

The distribution parameters may include a mixing coefficient, a mean, and a covariance. If the existing features exist in the form of feature vectors, the distribution parameters may include a mixing coefficient, a mean vector, and a covariance matrix, and the mixing coefficient may be a ratio of the number of samples belonging to the corresponding Gaussian distribution to the total number of samples. The training set may include a first training set with complete features and a second training set with missing features, and the existing features include the first existing feature, the second existing feature, and the first training set in the first training set. The third existing feature in the two training sets, the first existing feature corresponds to the third existing feature, the second existing feature corresponds to the missing feature, and step 202 may specifically include : Estimating an initial mixing coefficient, an initial mean value, and an initial agreement of each elementary Gaussian distribution of the preset mixed Gaussian model according to the first existing feature, the second existing feature, and the maximum likelihood estimation calculation variance.

It should be noted that the calculation of the maximum likelihood estimation may include:

Calculate the mixing coefficient of the i-th Gaussian distribution:

Calculate the mean of the i-th Gaussian distribution:

Calculate the covariance of the i-th Gaussian distribution:

γ _ij can be expressed as the probability that the sample x _j belongs to the i-th Gaussian distribution.

For example, the training set is D = {(x ₁ , y ₁ ), (x ₂ , y ₂ ), ..., (x _l , y _l ), x _{l + 1} , x _{l + 2} …, x _n }, features The complete first training set can be D ₁ = {(x ₁ , y ₁ ), (x ₂ , y ₂ ), ..., (x _l , y _l )}, and the second training set with missing features can be D ₂ = {x _{l + 1} , x _{l + 2} …, x _n }, the first existing feature in the first training set may be x _j , j = 1, ..., l, and the second existing feature may be y _i , i = 1, ..., l, the third existing feature in the second training set may be x _j , j = 1l + 1, ..., n, and the missing feature in the second training set may be y _i , I = l + 1, ..., n; specific x _j can correspond (average number of courses per month, work address latitude and longitude, number of Internet trading products in a single month), and y _i can correspond to x _j belonging to the i category of The probability γ _{ij of the} "Internet transaction level", the probability value of the class to which x _j belongs is 1, and the other classes are 0. Therefore, the first existing feature “x _j ” and the second existing feature “γ _ij ” can be substituted into the maximum likelihood estimation calculation to calculate the initial mixing coefficient, the initial mean, and the initial covariance of each elementary Gaussian distribution.

203. According to the initial distribution parameter and the expectation calculation of the maximum expectation algorithm, a Gaussian distribution to which the missing feature belongs is initially estimated.

For the embodiment of the present application, corresponding to step 202, the step 203 may specifically include: the initial mixing coefficient, the initial mean, the initial covariance, the third existing feature, and the expectation calculation. A Gaussian distribution to which the missing feature belongs is estimated. Specifically, the probability of the missing feature belonging to each meta-Gaussian distribution may be preliminarily estimated based on the initial mixing coefficient, the initial mean, the initial covariance, the third existing feature, and the expectation calculation; The probability of the distribution, a Gaussian distribution to which the missing feature belongs is initially estimated.

It should be noted that the expectation calculation may include:

Among them, m can be the total number of Gaussian distribution.

For example, following the example described in step 202, after calculating the initial parameters π _i , μ _i , Σ _i , x _j , j = l + 1, ..., n, π _i , μ _i , Σ _i , respectively Substitute into the above formula and calculate the probability that the "Internet transaction level within six months" belongs to each Gaussian distribution, that is, the probability γ _ij that belongs to each "Internet transaction level within six months"category; the Gaussian distribution with the highest probability can be determined as "six months Intra-Internet transaction level "Gaussian distribution.

204. Iteratively update the distribution parameters of each meta-Gaussian distribution according to the existing feature, the associated Gaussian distribution, and the maximum likelihood estimation calculation, and iteratively estimate the Gaussian distribution to which the missing feature belongs.

It should be noted that according to step 204, it is possible to iteratively update the distribution parameters of the respective Gaussian distributions according to all features, that is, the upper limit value in the maximum likelihood estimation calculation is changed from l to n:

Calculate the mixing coefficient of the i-th Gaussian distribution:

Calculate the mean of the i-th Gaussian distribution:

Calculate the covariance of the i-th Gaussian distribution:

205. When the iteratively updated distribution parameters converge, obtain an estimated Gaussian distribution based on the converged distribution parameters and the expectation, and estimate a value corresponding to the missing feature according to the estimated Gaussian distribution.

For the embodiment of the present application, in order to determine whether the updated distribution parameter converges, the method may further include: calculating a parameter difference value of the distribution parameter updated in two iterations before and after; if the parameter difference value is less than a preset threshold, determining update The distribution parameters converge.

Specifically, if the distribution parameters are the mixing coefficient, the mean, and the covariance, the difference of the mixing coefficient updated two times before and after iteration may be calculated; if the difference of the mixing coefficient is less than a preset mixing coefficient threshold, the convergence of the mixing coefficient may be determined. Or calculate the difference between the mean values updated before and after two iterations; if the mean difference is less than a preset mean value threshold, then the mean value can be determined to converge. Or calculate the difference between the updated covariances before and after two iterations; if the covariance difference is less than a preset covariance threshold, the covariance convergence can be determined. That is, when any one of the above parameters converges, the iterative maximum likelihood estimation calculation and the expectation calculation can be stopped. At this time, the missing feature is estimated according to the Gaussian distribution to which the missing feature belongs in the last iteration estimation. The preset mixing coefficient threshold, the preset average threshold, and the preset covariance threshold may all be set according to user requirements, or may be set according to a system default mode, which are not limited in this embodiment of the present application. .

206: Fill the values corresponding to the missing features into the crowd performance prediction training set, and train the crowd performance prediction model according to the crowd performance prediction training set after filling the values corresponding to the missing features.

It should be noted that, according to the positions of existing features, the values corresponding to missing features can be determined to be filled in the position of the crowd performance prediction training set. If the training set before filling the values corresponding to missing features is D = {(x ₁ , y ₁ ), (x ₂ , y ₂ ), ..., (x _l , y _l ), x _{l + 1} , x _{l + 2} …, x _n ), then you can use x _{l + 1} , x _{l + 2 respectively} The position of…, x _n is filled with y _{l + 1} , y _{l + 2} …, y _n , and finally y _{l + 1} , y _{l + 2} …, y _{n is} missing. The training set can be D = {(x ₁ , y ₁ ), (x ₂ , y ₂ ), ..., (x _l , y _l ), (x _{l + 1} , y _{l + 1} ), (x _{l + 2} , y _{l + 1} ) ..., (x _n , y _n )}. In addition, in the embodiment of the present application, the crowd performance prediction model may be a decision tree model or a logistic regression model, etc., for determining a crowd performance level. Specifically, if the crowd performance prediction model is a decision tree model, a decision tree algorithm can be used to train a crowd performance prediction training set after filling in missing features to obtain a decision tree model. If the crowd performance prediction model is a logistic regression model model, the logistic regression model algorithm can be used to train the crowd performance prediction training set after filling the missing features to obtain a logistic regression model model.

In order to better understand the embodiments of the present application, the following application scenarios are provided, including but not limited to this: Assume that the training set for predicting the performance characteristics of the crowd has training samples 1: (average number of monthly learning courses 100, Internet transaction level 1 within six months), Training sample 2 (average number of monthly courses 50, Internet transaction level 2 in half a year), training sample 3 (average number of monthly courses 60, Internet transaction level 2 in half a year), training sample 4 (average number of monthly courses 80,) , Training sample 5 (the average number of courses per month is 70,), the missing feature is "Internet trading levels within six months" in training samples 4 and 5. There are two categories of Internet transaction levels within six months, respectively: within six months Internet transaction level 1, Internet transaction level 2 within half a year.

First, the "average number of learning courses per month" and "internet trading levels within half a year" in training sample 1, training sample 2, and training sample 3 can be substituted into the maximum likelihood estimation calculation to obtain preliminary estimates of π _i and μ, respectively. _i , ∑ _i , and then substitute the "average number of learning courses per month" of training sample 4 and training sample 5 into the expectation calculation, and calculate the "Internet transaction level within six months" in training sample 4 as Internet transaction level 1 within six months respectively The probability of belonging to Internet transaction level 2 within half a year. If the probability of Internet transaction level 1 within half a year is greater than the probability of Internet transaction level 2 within half a year, the “Internet transaction level within six months” in training sample 4 can be determined The categories are: Level 1 of Internet transactions within six months. In the same way, the category of “Internet transaction level within six months” in training sample 5 can be calculated: Internet transaction level 1 within six months.

Then, you can substitute training sample 1, training sample 2, training sample 3, training sample 4 and training sample 5 "average learning courses per month" and "internet trading level within half a year" into the maximum likelihood estimation calculation, iteratively updated π _i, μ _i, Σ _i, and are updated training samples 4 and the training sample 5 in the category "within six months of Internet transaction level" belongs according to the updated π _i, μ _i, Σ _i and expectations calculated until π _i "within six months of Internet transaction level" category, μ _i, when Σ _i convergence, the estimated belongs as the final estimation results, such as finally determined: the training sample 4 "within six months of Internet transactions grade" category belongs: six months Intra-Internet Trading Level 1; The category of "Internet Trading Level in Six Months" in Training Sample 5 is: Internet Trading Level 2 in Six Months.

Therefore, after filling in the values corresponding to the missing features, the training set obtained can be: training sample 1: (average number of monthly learning courses 100, Internet transaction level 1 in half a year), training sample 2 (average number of monthly learning courses 50, half a year Internet transaction level 2), training sample 3 (average number of monthly courses 60, Internet transaction level 2 in half a year), training sample 4 (average number of monthly courses 80, Internet transaction level 1 in half a year), training sample 5 (monthly The average number of courses studied is 70, and the level of Internet transactions in half a year is 2).

Another method for processing missing features in the prediction of population performance characteristics provided by the embodiments of the present application is in line with the traditional filling methods such as special value interpolation, hot card interpolation, etc., which are currently used to fill missing characteristics in the crowd performance prediction training set. In contrast, the embodiment of the present application can obtain a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set, and the preset mixed Gaussian model is composed of a multivariate Gaussian distribution corresponding to the missing features; The maximum expectation algorithm of the existing feature and the preset mixed Gaussian model is described, and the missing feature is estimated. At the same time, the estimated missing features can be filled into the crowd performance prediction training set, so that the missing data can be filled based on a preset mixed Gaussian model corresponding to the missing features. The composition of the multivariate Gaussian distribution corresponding to the missing features can ensure that the missing features after filling reflect their distribution status and are related to themselves. It can avoid the bias of missing feature filling and can avoid the prediction results of the trained prediction model and the corresponding features. There is a bias in the association, which can improve the prediction accuracy of the trained prediction model.

Further, as a specific implementation of FIG. 1, an embodiment of the present application provides a missing feature processing device in crowd performance prediction. As shown in FIG. 3, the device includes: an obtaining unit 31, an estimation unit 32, and a filling unit 33. .

The obtaining unit 31 may be configured to obtain a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set, where the preset mixed Gaussian model is composed of a multivariate Gaussian distribution corresponding to the missing features. The obtaining unit 31 is a main functional module for obtaining a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set in the device.

The estimation unit 32 may be configured to estimate a value corresponding to the missing feature according to a maximum expected algorithm of the existing feature and the preset mixed Gaussian model. The estimation unit 32 is a main function module for estimating the missing feature according to the maximum expectation algorithm of the existing feature and the preset mixed Gaussian model, and is also a core module.

The filling unit 33 may be configured to fill a value corresponding to the missing feature into the crowd performance prediction training set. The filling unit 33 is a main functional module in the device that fills the values corresponding to the missing features into the crowd performance prediction training set.

For the embodiment of the present application, the maximum expected algorithm may include a maximum likelihood estimation calculation and an expected calculation, and the estimation unit 32 may include an estimation module 321, an update module 322, and an acquisition module 323, as shown in FIG. 4.

The estimation module 321 may be configured to estimate an initial distribution parameter of each meta-Gaussian distribution of the preset mixed Gaussian model according to the existing features and the maximum likelihood estimation calculation.

The estimation module 321 may be further configured to initially estimate a Gaussian distribution to which the missing feature belongs according to the initial distribution parameter and the expectation calculation.

The updating module 322 may be configured to iteratively update the distribution parameters of the respective Gaussian distributions according to the existing features, the associated Gaussian distribution, and the maximum likelihood estimation calculation, and iteratively update the missing Gaussian distribution to which the feature belongs.

The obtaining module 323 may be configured to obtain a Gaussian distribution estimated according to the converged distribution parameters and the expectation when the iteratively updated distribution parameters converge.

The estimation module 321 may be further configured to estimate a value corresponding to the missing feature according to a last estimated Gaussian distribution.

In a specific application scenario, the estimation module 321 may be specifically configured to initially estimate a probability that the missing feature belongs to each meta-Gaussian distribution according to the initial distribution parameter and the expectation calculation; and according to the belonging to each meta-Gaussian distribution The probability of the distribution, a Gaussian distribution to which the missing feature belongs is initially estimated.

It should be noted that the training set includes a first training set with complete features and a second training set with missing features, and the existing features include the first existing feature and the second existing feature in the first training set. And the third existing feature in the second training set, the first existing feature corresponds to the third existing feature, the second existing feature corresponds to the missing feature, and the estimation Module 321 may be specifically configured to estimate an initial mixing coefficient of each elementary Gaussian distribution of the preset mixed Gaussian model according to the first existing feature, the second existing feature, and the maximum likelihood estimation calculation. , Initial mean and initial covariance.

The estimation module 321 may also be specifically used for the initial mixing coefficient, the initial mean, the initial covariance, the third existing feature, and the expectation calculation to initially estimate the Gaussian distribution to which the missing feature belongs.

For the embodiment of the present application, in order to determine whether the updated distribution parameters converge, the estimation unit 32 may further include a calculation module 324 and a determination module 325.

The calculation module 324 may be configured to calculate a parameter difference of the distribution parameter updated by two iterations before and after.

The determining module 325 may be configured to determine that the updated distribution parameter converges if the parameter difference is less than a preset threshold.

For the embodiment of the present application, in order to obtain a preset mixed Gaussian model corresponding to a missing feature, the apparatus may further include a determining unit 34 and a constructing unit 35.

The determining unit 34 may be configured to determine a multivariate Gaussian distribution corresponding to the missing feature. The determining unit is a main functional module for determining a multivariate Gaussian distribution corresponding to the missing feature in the device.

The constructing unit 35 may be configured to construct the preset mixed Gaussian model according to the multivariate Gaussian distribution. The constructing unit 35 is a main functional module for constructing the preset mixed Gaussian model according to the multivariate Gaussian distribution in the device.

In addition, in order to obtain a crowd performance prediction model, the device may further include a training unit 36.

The training unit 36 may be configured to train a crowd performance prediction model according to a crowd performance prediction training set after filling in missing features. The training unit 36 is a main functional module for training a crowd performance prediction model in the device according to a crowd performance prediction training set after filling in missing features.

It should be noted that for other corresponding descriptions of the functional modules involved in the missing feature processing device in a crowd performance prediction provided in the embodiments of the present application, reference may be made to the corresponding description of the method shown in FIG. 1, and details are not described herein again.

Based on the above method shown in FIG. 1, correspondingly, an embodiment of the present application further provides a computer non-volatile readable storage medium that stores computer-readable instructions. When the computer-readable instructions are executed by a processor, The following steps are implemented: acquiring existing features and preset mixed Gaussian models corresponding to missing features in a crowd performance prediction training set, the preset mixed Gaussian models consisting of a multivariate Gaussian distribution corresponding to the missing features; and according to the existing features And a maximum expectation algorithm of the preset mixed Gaussian model to estimate a value corresponding to the missing feature; and filling the value corresponding to the missing feature to the crowd performance prediction training set.

Based on the embodiment of the method shown in FIG. 1 and the device shown in FIG. 3, the embodiment of the present application further provides a physical structure diagram of a computer device. As shown in FIG. 5, the computer device includes a processor 41, The memory 42 and computer-readable instructions stored on the memory 42 and executable on the processor, where the memory 42 and the processor 41 are both arranged on the bus 43 and the processor 41 implements the following when the computer-readable instructions are executed Step: Obtain a preset mixed Gaussian model corresponding to the existing features and missing features in the crowd performance prediction training set, the preset mixed Gaussian model is composed of a multivariate Gaussian distribution corresponding to the missing features; The maximum expectation algorithm of the preset mixed Gaussian model is used to estimate the value corresponding to the missing feature; the value corresponding to the missing feature is filled into the crowd performance prediction training set.

Through the technical solution of the present application, a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set can be obtained, and the preset mixed Gaussian model is composed of a multivariate Gaussian distribution corresponding to the missing features; The maximum expectation algorithm of the existing feature and the preset mixed Gaussian model estimates a value corresponding to the missing feature. At the same time, the values corresponding to the estimated missing features can be filled into the crowd performance prediction training set, so that the missing data can be filled based on the preset mixed Gaussian model corresponding to the missing features. Since the preset mixed Gaussian model is It consists of a multivariate Gaussian distribution corresponding to the missing feature, so it can ensure that the missing feature after filling reflects its distribution state and is related to itself, can avoid the bias of missing feature filling, and can avoid the prediction result and corresponding of the trained prediction model. There is a bias in the association between features, which can improve the prediction accuracy of the trained prediction model.

Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present application may be implemented by a general-purpose computing device, and they may be concentrated on a single computing device or distributed in a network composed of multiple computing devices. Above, optionally, they can be implemented with computer-readable instruction code executable by the computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, can be different from this The steps shown or described are performed in sequence, either by making them into individual integrated circuit modules, or by making multiple modules or steps into a single integrated circuit module. As such, this application is not limited to any particular combination of hardware and software.

The above description is only a preferred embodiment of the present application, and is not intended to limit the present application. For those skilled in the art, this application may have various modifications and changes. Any modification, equivalent replacement, or improvement made within the spirit and principle of this application shall be included in the protection scope of this application.

Claims (20)

A method for processing missing features in crowd performance prediction, which is characterized by: Obtaining a preset mixed Gaussian model corresponding to existing features and missing features in a crowd performance prediction training set, the preset mixed Gaussian model consisting of a multivariate Gaussian distribution corresponding to the missing features; Estimating a value corresponding to the missing feature according to a maximum expectation algorithm of the existing feature and the preset mixed Gaussian model; The values corresponding to the missing features are filled into the crowd performance prediction training set. The method according to claim 1, wherein the maximum expectation algorithm comprises a maximum likelihood estimation calculation and an expectation calculation, and estimates the maximum expected algorithm based on the existing feature and the maximum expected algorithm of the preset mixed Gaussian model. Describe the values corresponding to missing features, including: Estimating an initial distribution parameter of each elementary Gaussian distribution of the preset mixed Gaussian model according to the existing feature and the maximum likelihood estimation calculation; Preliminary estimate a Gaussian distribution to which the missing feature belongs according to the initial distribution parameter and the expectation calculation; Calculating the distribution parameters of the respective Gaussian distribution iteratively according to the existing feature, the associated Gaussian distribution, and the maximum likelihood estimation calculation, and iteratively updating the Gaussian distribution to which the missing feature belongs; When the iteratively updated distribution parameters converge, an estimated Gaussian distribution is calculated according to the converged distribution parameters and the expectation, and a value corresponding to the missing feature is estimated according to the last estimated Gaussian distribution. The method according to claim 2, wherein the preliminary estimation of the Gaussian distribution to which the missing feature belongs according to the initial distribution parameter and the expectation calculation comprises: Preliminarily estimate the probability that the missing feature belongs to each elementary Gaussian distribution according to the initial distribution parameter and the expectation calculation; According to the probabilities belonging to each meta-Gaussian distribution, a Gaussian distribution to which the missing feature belongs is initially estimated. The method according to claim 2, wherein the training set includes a first training set with complete features and a second training set with missing features, and the existing features include the first training set in the first training set. There are a feature, a second existing feature, and a third existing feature in the second training set, the first existing feature corresponds to the third existing feature, and the second existing feature corresponds to the The missing features correspond; the estimating the initial distribution parameters of each elementary Gaussian distribution of the preset mixed Gaussian model based on the existing features and the maximum likelihood estimation calculation includes: Estimate the initial mixing coefficient, the initial mean, and the initial covariance of each meta-Gaussian distribution of the preset mixed Gaussian model according to the first existing feature, the second existing feature, and the maximum likelihood estimation calculation ; The preliminary estimation of the Gaussian distribution to which the missing feature belongs according to the initial distribution parameter and the expectation calculation includes: Based on the initial mixing coefficient, the initial mean, the initial covariance, the third existing feature, and the expectation calculation, a Gaussian distribution to which the missing feature belongs is initially estimated. The method according to claim 2, wherein before the iteratively updated distribution parameter converges, before obtaining the estimated Gaussian distribution based on the converged distribution parameter and the expectation calculation, the method further comprises: Calculate the parameter difference of the distribution parameters updated before and after two iterations; If the parameter difference is less than a preset threshold, it is determined that the updated distribution parameter converges. The method according to claim 1, wherein before the obtaining a preset mixed Gaussian model corresponding to an existing feature and a missing feature in a crowd performance prediction training set, the method further comprises: Determining a multivariate Gaussian distribution corresponding to the missing feature; Constructing the preset mixed Gaussian model according to the multivariate Gaussian distribution. The method according to claim 1, wherein after the method fills the missing feature into the crowd performance prediction training set, the method further comprises: The population performance prediction model is trained according to the population performance prediction training set after filling the missing features. A missing feature processing device in crowd performance prediction is characterized in that it includes: An obtaining unit, configured to obtain an existing feature and a preset mixed Gaussian model corresponding to the missing feature in the training set for predicting the performance of the crowd; the preset mixed Gaussian model consists of a multivariate Gaussian distribution corresponding to the missing feature; An estimation unit, configured to estimate a value corresponding to the missing feature according to a maximum expectation algorithm of the existing feature and the preset mixed Gaussian model; The filling unit is configured to fill a value corresponding to the missing feature into the crowd performance prediction training set. The apparatus according to claim 8, wherein the maximum expectation algorithm comprises a maximum likelihood estimation calculation and an expectation calculation, and the estimation unit comprises: An estimation module, configured to estimate, based on the existing features and the maximum likelihood estimation calculation, an initial distribution parameter of each meta-Gaussian distribution of the preset mixed Gaussian model; The estimation module is further configured to initially estimate a Gaussian distribution to which the missing feature belongs according to the initial distribution parameter and the expectation calculation; An update module, configured to iteratively update the distribution parameters of the respective Gaussian distributions according to the existing features, the associated Gaussian distribution, and the maximum likelihood estimation calculation, and iteratively update the Gaussian to which the missing features belong. distributed; An acquisition module, configured to obtain a Gaussian distribution estimated to be calculated according to the converged distribution parameters and the expectation when the iteratively updated distribution parameters converge; The estimation module may be further configured to estimate a value corresponding to the missing feature according to a last estimated Gaussian distribution. The apparatus according to claim 9, wherein the estimation module is specifically configured to initially estimate the probability that the missing feature belongs to each elementary Gaussian distribution according to the initial distribution parameter and the expectation calculation; The probability of each Gaussian distribution is described, and the Gaussian distribution to which the missing feature belongs is estimated. The apparatus according to claim 9, wherein the training set includes a first training set with complete features and a second training set with missing features, and the existing features include the first training set in the first training set. There are a feature, a second existing feature, and a third existing feature in the second training set, the first existing feature corresponds to the third existing feature, and the second existing feature corresponds to the The missing feature corresponds; the estimation module is specifically configured to estimate each element of the preset hybrid Gaussian model based on the first existing feature, the second existing feature, and the maximum likelihood estimation calculation. The initial mixing coefficient, initial mean, and initial covariance of the Gaussian distribution. The estimation module is further configured to initially estimate a Gaussian distribution to which the missing feature belongs according to the initial mixing coefficient, the initial mean, the initial covariance, the third existing feature, and the expectation calculation. The apparatus according to claim 9, wherein the estimation unit comprises: A calculation module for calculating parameter differences of the distribution parameters updated in two iterations before and after; A determining module, configured to determine that the updated distribution parameter converges if the parameter difference is less than a preset threshold. The apparatus according to claim 8, further comprising: A determining unit, configured to determine a multivariate Gaussian distribution corresponding to the missing feature; A construction unit is configured to construct the preset mixed Gaussian model according to the multivariate Gaussian distribution. The apparatus according to claim 13, further comprising: A training unit is configured to train a crowd performance prediction model according to a crowd performance prediction training set after filling in the values corresponding to missing features. A computer non-volatile readable storage medium having computer readable instructions stored thereon, characterized in that when the computer readable instructions are executed by a processor, a method for processing missing features in a crowd performance prediction includes: Acquire the existing features and the preset mixed Gaussian models corresponding to the missing features in the crowd performance prediction training set, the preset mixed Gaussian models are composed of a multivariate Gaussian distribution corresponding to the missing features; according to the existing features and the predictive A maximum expectation algorithm of the mixed Gaussian model is used to estimate the value corresponding to the missing feature; the value corresponding to the missing feature is filled into the crowd performance prediction training set. The computer non-volatile readable storage medium according to claim 15, wherein the maximum expectation algorithm comprises a maximum likelihood estimation calculation and an expectation calculation, and the computer-readable instructions implement all operations when executed by a processor. The method of estimating a value corresponding to the missing feature according to the maximum expectation algorithm of the existing feature and the preset mixed Gaussian model includes: estimating the value based on the existing feature and the maximum likelihood estimation calculation. Preset the initial distribution parameters of each Gaussian distribution of the mixed Gaussian model; based on the initial distribution parameters and the expectation calculation, preliminary estimate the Gaussian distribution to which the missing feature belongs; Calculation of the distribution and the maximum likelihood estimation, iteratively update the distribution parameters of the respective Gaussian distributions, and iteratively update the Gaussian distribution to which the missing feature belongs; when the iteratively updated distribution parameters converge, obtain the distribution parameters according to the convergence Calculate the estimated Gaussian distribution with the expectation, and estimate the value corresponding to the missing feature according to the last estimated Gaussian distribution The computer non-volatile readable storage medium according to claim 15, wherein implementing the method when the processor executes the computer-readable instructions further comprises: filling a population corresponding to a value corresponding to the missing feature. The performance prediction training set trains a crowd performance prediction model. A computer device includes a memory, a processor, and computer-readable instructions stored on the memory and executable on the processor, characterized in that, when the processor executes the computer-readable instructions, it realizes Missing feature processing methods, including: Acquire the existing features and the preset mixed Gaussian models corresponding to the missing features in the crowd performance prediction training set, the preset mixed Gaussian models are composed of a multivariate Gaussian distribution corresponding to the missing features; according to the existing features and the predictive A maximum expectation algorithm of the mixed Gaussian model is used to estimate the value corresponding to the missing feature; the value corresponding to the missing feature is filled into the crowd performance prediction training set. The computer device according to claim 18, wherein the maximum expectation algorithm comprises a maximum likelihood estimation calculation and an expectation calculation, and the computer-readable instructions implement the according to the existing features when executed by a processor. And a maximum expectation algorithm of the preset mixed Gaussian model to estimate a value corresponding to the missing feature includes: estimating each of the preset mixed Gaussian model according to the existing feature and the maximum likelihood estimation calculation. An initial distribution parameter of the meta-Gaussian distribution; a preliminary estimation of the Gaussian distribution to which the missing feature belongs according to the initial distribution parameter and the expectation calculation; according to the existing feature, the belonging Gaussian distribution, and the maximum likelihood For the estimation calculation, iteratively updates the distribution parameters of the Gaussian distributions of each element, and iteratively updates the Gaussian distribution to which the missing feature belongs; when the iteratively updated distribution parameters converge, obtaining the estimated Gaussian distribution, and the value corresponding to the missing feature is estimated according to the last estimated Gaussian distribution. The computer device according to claim 18, wherein implementing the method when the processor executes the computer-readable instructions further comprises: training a crowd performance based on a crowd performance prediction training set after filling in values corresponding to missing features. Forecasting model.

Images