Systematic Analysis of Predictive Modeling Methods in Stock Markets

Abstract

Stock Market Prediction is a challenging task due to the volatile, unpredictable and chaotic nature of the stock market. Global digitization has revamped SMP and trading techniques. Many researchers have employed Machine learning for predicting future value of stocks helping investors to make safe and wise financial decisions. This study systematically examines the traditional prediction methods and the modern approaches that utilize Artificial Intelligence and Machine Learning for the task of prediction. The study compares and contrasts various supervised and unsupervised techniques and Artificial Neural Networks that use temporal data for prediction. Performance of algorithms depends on the dynamic input data, and the nature of forecast. Data fitting is an important concern for identifying, analyzing and predicting future instances. Extensive research is required to build appropriate modules for data pre-processing, analysis, and prediction. Comparing the performance of ML algorithms with traditional methods is required to prove their effectiveness. The study explores the strengths of various ML algorithms to develop a basic understanding, and paves the way for further research in the field of Stock Market Prediction.