Performance Comparison of Back propagation Neural Network and Extreme Learning machine for Multinomial Classification Task

Abstract

Classification and prediction tasks continue to play a vital role in the area of computer science and data processing. Clustering and classification in Data Mining are used in various domains to give meaning to the available data. Data Mining has especially become popular in the fields of forensic science, fraud analysis and healthcare, as it reduces costs in time and money. In classification modeling the data is classified to make predictions about new data. Using old data to predict new data has the danger of being too fitted on the old data. But that problem can be solved by using soft computing tools which generalizes the same type of data into one class and rest to the other which are known as binary classifiers. This paper describes and compares the application of two popular machine learning methods: Back propagation neural network and Extreme learning machine which are used as multiclass classifiers. These two approaches are applied on same type of multi class classification datasets and the work tries to generate some comparative inferences from training and testing results. The datasets are taken from UCI learning repository.