Boosting as a dimensionality reduction tool for audio classification

In this paper we present a modified AdaBoost algorithm that can be used for dimensionality reduction in audio classification problems. The algorithm is modified to work as a feature selector for a four way classification problem. It is compared with principal component analysis (PCA), which is a popular tool for reducing the dimensions of high-dimensional data without losing significant scatter information. Both algorithms are applied to a four way audio classification problem and the results are presented.