Evaluation of Eigen Value Extraction and Automatic Computational Classification of Corpus in Scene Scheduling of Sports and Health IoT Platform

Abstract

This study constructed a corpus by collecting samples of applications in exercise health information. Based on the extracted 8 eigen values, this study used the Fisher score algorithm to score the eigen values and then formed a combination of 8 groups of eigen values according to the score ranking. This study performed computerized automatic classification using cross-checking based on three methods, support vector machines, neural networks, and Naive Bayes model. By analyzing the experimental results, the optimal eigen value combination and the optimal classification algorithm were finally derived in this study.