这个程序实现了Francis R. Bach的Bolasso算法，用于特征选取和预测。主要用于高纬度问题的特征选取，它使用了带有Bootstrap方法的自助抽样的正则化回归，并使用了Karl Skoglund的lars实现。-This procedure achieved Francis R. Bach s Bolasso algorithms for feature selection and forecasting. The main problem for high-latitude feature selection, it uses a method of self-help Bootstrap sampling Tikhonov reunification, and Karl Skoglund used to achieve the lars.