山西大学学报（自然科学版）2017，Vol.40Issue(4)：750-755,6.DOI:10.13451/j.cnki.shanxi.univ(nat.sci.).2017.04.010
一种基于混合度的层次粒度SVM算法
A Hierarchical Granular Support Vector Machine Algorithm based on Mixed
摘要
Abstract
The increasing size and dimensionality of real-world datasets make it necessary to design efficient classification algorithms.Support vector machine(SVM) is an acknowledged classification algorithm with good performance,and some SVM algorithms are aimed at reducing the number of support vector to improve the efficiency of classification.This paper proposes a hierarchical granularity support vector machine algorithm based on mixed,namely MHG-SVM,using the mixing degree to improve the existing hierarchical granularity support vector machine algorithm.The proposed algorithm defined a data granularity of confidence and a parameter to select important classification information,and then put them in training set to SVM training.The experimental results indicate that the improved algorithm is suitable for processing large number of observations and can effectively accelerate SVM learning while keeping the classification precision.关键词
支持向量/分类精度/混合度/置信度/大规模数据集Key words
support vector/classification precision/mixing degree/granularity of confidence/large number of observations分类
信息技术与安全科学引用本文复制引用
程凤伟,郭虎升..一种基于混合度的层次粒度SVM算法[J].山西大学学报（自然科学版）,2017,40(4):750-755,6.基金项目
山西省自然科学基金(2015021096),山西省高等学校科技创新项目(2015110) （2015021096）
