Efficient Methods to Improve the Performance of Supervised Learning Models

Abstract

Supervised Learning is defined as training a model with input data that includes the result itself. There are large number of supervised learning algorithms and great number of models. Each model has its own merits and demerits and performs differently. There are many data preprocessing techniques and hence the combination of several data preprocessing techniques can increase the performance of the present supervised learning models. Primary data can not be fed directly to the learning model because it can hold a lot of noise. It needs to be preprocessed using various data preprocessing techniques. We have analysed and compared different data preprocessing techniques and their combinations. Comparison is done using various performance metrics and the combination of different data preprocessing is applied to different model. We have done categorical data handling, missing value treatment, feature scaling and feature extraction as the data preprocessing steps. Through the comparison, we found which technique is better for which type of models. We used the California census data for our study.