A Comparative Analysis of Machine Learning Algorithms for Classification Purpose
Vraj Sheth, Urvashi Tripathi, Ankit Sharma
Abstract
A few of the popular data-mining techniques are clustering, classification, and association. The classification process simplifies the process of identifying and accessing data. Classification of data is crucial for risk management, compliance, and data security. Classifying data facilitates its search-ability and traceability by categorising the information. Each data mining model has a distinct level of information. The success of a model is solely determined by the datasets being used, as there is no such thing as an excellent or a poor model. As a part of this study, we examine how accurate different classification algorithms are on diverse datasets. On five different datasets, four classification models are compared: Decision tree, SVM, Naive Bayesian, and K-nearest neighbor. The Naive Bayesian algorithm is proven to be the most effective among other algorithms.