Python 第三方模块 机器学习 Scikit-Learn模块 无监督学习2 协方差估计,流形学习,高斯混合模型

一.covariance
1.简介:

该模块用于对协方差进行估计

2.使用
(1)类:

"最大似然协方差估计量"(Maximum likelihood covariance estimator):class sklearn.covariance.EmpiricalCovariance([store_precision=True,assume_centered=False])
用于检测"高斯分布数据集"(Gaussian distributed dataset)中"离群值"(outliers)的对象:class sklearn.covariance.EllipticEnvelope([store_precision=True,assume_centered=False,support_fraction=None,contamination=0.1,random_state=None])
带有"l1-惩罚估计量"(l1-penalized estimator)的"稀疏逆协方差"(Sparse inverse covariance)估计:class sklearn.covariance.GraphicalLasso([alpha=0.01,mode='cd',tol=0.0001,enet_tol=0.0001,max_iter=100,verbose=False,assume_centered=False])
带有"l1惩罚"(l1 penalty)的"交叉验证选择"(cross-validated choice):class sklearn.covariance.GraphicalLassoCV([alphas=4,n_refinements=4,cv=None,tol=0.0001,enet_tol=0.0001,max_iter=100,mode='cd',n_jobs=None,verbose=False,assume_centered=False])
"Ledoit-Wolf估计量"(Ledoit-Wolf Estimator):class sklearn.covariance.LedoitWolf([store_precision=True,assume_centered=False,block_size=1000])
"最小协方差决定因素"(Minimum Covariance Determinant;MCD):class sklearn.covariance.MinCovDet([store_precision=