Boosting理论基础

The Theory Base of Boosting

Boosting是提高学习算法准确度的有效方法。本文主要介绍了Boosting的问题框架PAC模型、与Boosting相似并有助于AdaBoost研完的在线分配模型和AdaBoost算法,并对AdaBoost算法的参数和弱假设选择等进行了分析。

Boosting is an effective method for improving thd accuracy of any given learning algorithm,which generate multiple versions of a hypothesis and combine them to create an aggregate hypothesis,This paper first introduces the problem framework of Boostinig:PAC learning model,and Online prediction model,a similar algorithm with boosting but with great help for boosting's research. Besides,if also introduces and analysis the algorithm of adaboost itself,as well as its parameters and week hypotheses'selection.