A SYSTEMATIC LITERATURE REVIEW ON SUPERVISED MACHINE LEARNING ALGORITHMS

Abstract

There are many researchers and data analyst in large companies around the world applied Machine Learning (ML) in the various study. ML is a subset of Artificial Intelligence (AI) which play a significant role in analyzing the big data. In general, the Supervised Machine Learning (SML), one type of ML, generates the desired output and makes a prediction based on the trained dataset provided in the input. Various algorithms under SML including Naïve Bayes, Logistic Regression, Random Forest, J48, CART, Multi-Layer Perceptron, Support Vector Machine (SVM) which are common and famously used by researchers. There is a total of 305 studies that is being compiled initially in this paper reviews on SML classification algorithms by adopting Systematic Literature Review (SLR) method. After sorting the papers according to the selection criteria and data extraction, 61 final studies were selected. As a conclusion, SML had been mostly used in classifying spam and text and also in healthcare and medically related classification research. It is also founded that SVM and Artificial Neural Network (ANN) are the top two performing algorithms in classification. In the future work, it is recommended to expand the study to include more assessment measures of SML algorithms and unsupervised machine learning (UML).