A Study For Ensemble Learning Based On SVM

Support Vector Machine based on statistical learning theory and structural risk minimization is a new learning method, it has good generalization performance. By training and combining some accurate and diverse classifiers,ensemble learning provides a novel approach for improving the generalization performance of classification systems.This article will introduce ensemble learning technology into support vector machine learning for improving the generalization ability of support vector machine. The main research work and innovative results are as follows:(1) Research on the support vector machines and ensemble learning theory, implementation and practical application,and introduce two classic ensemble learning algorithms:Bagging and Boosting.(2) A good ensemble is one where the base classifiers in the ensemble are both accurate and tend to err in different parts of the instance space.Feature selection can improve classifiers accuracy and diverse, disturb support vector machine model parameters, can also increase the diverse of individual classifiers. This paper introduces Relief filter feature selection algorithm and the prediction of risk embedded feature selection methods involved in ensemble learning and randomly selected SVM model parameters in low-bias region, proposed two based on feature selection and low-bias SVM ensemble learning algorithm.(3) A new selective ensemble of support vector machine learning algorithm. Propose a continuously remove the worst classifier method, and use the logistic function transform the decision value of SVM classifiers, improve the generalization performance of the algorithm.(4) Validate the three proposed algorithms on the UCI datasets, and achieved expected results.