用于特征降维，特征融合，相关分析等,正确率可以达到98％，包括主成分分析、因子分析、贝叶斯分析，采用了小波去噪的思想，包括回归分析和概率统计，gmcalab 快速广义的形态分量分析，分析了该信号的时域、频域、倒谱，循环谱等。-For feature reduction, feature fusion, correlation analysis, Accuracy can reach 98 , Including principal component analysis, factor analysis, Bayesian analysis, Using wavelet denoising thought, Including regression analysis and probability and statistics, gmcalab fast generalized form component analysis, Analysis of the signal time domain, frequency domain, cepstrum, cyclic spectrum, etc.