An ensemble learning algorithm based on Lasso selection

Abstract

Ensemble learning, especially selective ensemble learning is now becoming more and more popular in the field of machine learning. This paper introduces a new ensemble algorithm, named Lasso-Bagging Trees ensemble algorithm. This algorithm is in order to improve the whole learning ability, which is a combination of tree predictors and this method chooses and ensembles trees based on the shrinkage estimation of lasso technology. Compared with a series of other learning algorithms, it demonstrates better generalization ability and higher efficiency.