Ice hockey competition condition prediction method based on feature selection and parameter optimization

Abstract

An ice hockey competition condition prediction method based on feature selection and parameter optimization is characterized by comprising: firstly, using an L1 norm feature retention rule for removing features with large sparse scores and small correlation in an ice hockey data set, and achieving feature selection; secondly, aiming at the problems that a penalty factor C and a kernel function parameter g in an ice hockey competition condition prediction model based on a support vector machine have large influence on generalization ability and parameters are difficult to adjust, adopting a mixed GAPSO parameter optimization algorithm to optimize the penalty factor C and the kernel function parameter g of the support vector machine; and finally, predicting the ice hockey competition condition by adopting a K-fold cross validation method of a support vector machine. The method improves the operation speed and efficiency of the ice hockey game prediction model through the feature selection algorithm, and improves the ice hockey game condition prediction accuracy.

Description

technical field [0001] The invention belongs to big data processing technology, and in particular relates to a ice hockey game structure prediction technology, in particular to a ice hockey game situation prediction method based on feature selection and parameter optimization. Background technique [0002] With the development of information technology, the data dimension and data volume of competitive sports data are increasing exponentially. In ice hockey, the offensive position is also the defensive position. Compared with other sports, it emphasizes personnel cooperation and division of labor. The tactical layout of the offensive position and the transition between offense and defense all have a crucial impact on the game, so the effectiveness of surrounding data and the amount of data for comparison specimens have a direct impact on the prediction of the game; accurate prediction of the game situation can make the team avoid risks , learn from each other's strengths, w...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the invention is to solve the problem that the amount of ice hockey game data is quite large and the data dimension is high, the calculation is complicated, and the penalty factor C and the kernel function parameter g of the prediction model based on support vector machine have a great influence on its generalization ability, and the parameters are difficult to adjust , Invented a ice hockey prediction method based on feature selection and parameter optimization

Images