(The) performance analysis of bagging and boosting

Abstract

A complex computational task is solved by dividing it into a number of computationally simple tasks and then boosting the solutions to those tasks. The combination of experts is said to constitute a committee machine. They may be classified into static structures and dynamic structures. Boosting works by repeatedly running a given weak learning algorithm on various distributions over the training data, and then combining the classifiers produced by the weak learner into a single composite classifier. AdaaBoost is the most popular boosting algorithm. Bagging predictor is a method for generating multiple versions of a predictor and using these to get an Baggregated predictor. The aggregation averages over the versions when predicting a class. The multiple version are formed by making bootstrap replicates of the learning set and using these as new learning sets. In this study, we repot results of applying both techniques to a system that learns decision tree and testing on a