重要参数的提取，进行逐步线性回归，用于特征降维，特征融合，相关分析等,用MATLAB实现的压缩传感，通过反复训练模板能有较高的识别率，用于信号特征提取、信号消噪。- Extract important parameters, Stepwise linear regression, For feature reduction, feature fusion, correlation analysis, Using MATLAB compressed sensing, Through repeated training rBEnDtwlate have higher recognition rate, For feature extraction, signal de-noising.