基于粒化-融合的海量高维数据特征选择算法  

Feature Selection Algorithm Based on Granulation-Fusion for Massive High-Dimension Data

作　　者：冀素琴[1] 石洪波[1] 吕亚丽[1] 郭珉[1] 

机构地区：[1]山西财经大学信息管理学院,太原030006

出　　处：《模式识别与人工智能》2016年第7期590-597,共8页Pattern Recognition and Artificial Intelligence

基　　金：国家自然科学基金项目(No.60873100);山西省自然科学基金项目(No.2014011022-2;2013011016-4);中国博士后科学基金面上项目(No.2016M591409)资助~~

摘　　要：基于粒计算视角,提出粒化-融合框架下的海量高维数据特征选择算法.运用BLB(Bag of Little Bootstrap)的思想,首先将原始海量数据集粒化为小规模数据子集(粒),然后在每个粒上构建多个自助子集的套索模型,实现粒特征选择,最后,各粒特征选择结果按权重融合、排序,得到原始数据集的有序特征选择结果.人工数据集和真实数据集上的实验表明文中算法对海量高维数据集进行特征选择的可行性和有效性.From a granular computing perspective, a feature selection algorithm based on granutatlon^tuslon Ior massive and high-dimension data is proposed. By applying bag of little Bootstrap （BLB）, the original massive dataset is granulated into small subsets of data （ granularity）, and then features are selected by constructing multiple least absolute shrinkage and selection operator （LASSO） models on each granularity. Finally, features selected on each granularity are fused with different weights, and feature selection results are obtained on original dataset through ordering. Experimental results on artificial datasets and real datasets show that the proposed algorithm is feasible and effective for massive high-dimension datasets.

关 键 词：海量高维数据 特征选择 粒计算 套索(LASSO) 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]