International Journal of Advance Research, Ideas and Innovations in Technology © 2021, www.IJARIIT.com All Rights Reserved Page |1625 ISSN: 2454-132X Impact Factor: 6.078 (Volume 7, Issue 3 - V7I3-1911) Available online at: https://www.ijariit.com Review on the applications of machine learning models for stock market predictions: A literature survey Sanjana Hemaraju sanjanahemaraju99@gmail.com RV College of Engineering, Bengaluru, Karnataka Monish Singhal monishsinghal16@gmail.com RV College of Engineering, Bengaluru, Karnataka Dr N. S. Narahari nsnarahari@gmail.com RV College of Engineering, Bengaluru, Karnataka ABSTRACT Financial stocks values are non-linear, volatile, and chaotic, making them one of the most challenging financial time series to predict. The incentive of financial gain has led many researchers and academia to devise methods to predict the stock market, despite copious uncertainty. Because of their ability to recognize complex patterns in several applications, machine learning models are extensively researched among the most recent methods. In this paper, Support Vector Machine, Artificial Neural Networks and Case-based Reasoning for stock market prediction is surveyed. This paper also reviews sentiment analysis to highlight the behavioral trends of the stock market and its investors with the advent of technology.A generalised modelling methodology for applying machine learning techniques to the stock market is proposed in this paper . Keywords: Artificial Neural Network, Case-Based Reasoning, Sentiment Analysis, Stock Market, Support Vector Machine 1. INTRODUCTION Significant research is devoted to stock market prediction, one of the most strenuous time series problems owing to the volatile, complex, dynamic and non-linear nature of the market. With the advent of technology and increased availability of information, investing in the stock market has become popular amongst the general public. The potential for financial gain drives the need for a methodical approach to predict the market and understand its trends and movements. Despite the Efficient Market Hypothesis (EMH) [1] and Random Walk Hypothesis [2], numerous techniques have been applied for stock market prediction over many years, stock markets follow chaotic patterns and are unreliable therefore demanding further research on the development of models and algorithms. Several indices have been formed to assess the relative value of the stocks traded within a market. Fundamental analysis and technical analysis are the two different stock market forecasting approaches widely considered. J. J. Murphy [3] defined technical analysis as “the study of market action, primarily through the use of charts, for the purpose of forecasting future price trend” and believes price, volume and open interest are the three predominant sources of information available. They suggest three ideologies upon which technical analysis is based, namely “market action discounts everything”, “price moves in trends”, and “history re peats itself” [3]. On the other hand, Fundamental Analysis is a way of determining the value of a stock by consid ering economic and financial elements including macroeconomic and microeconomic factors like the economics of supply and demand, and the company’s performance [4]. Some experts and traders apply both approaches, specifically fundamental analysis to determi ne what stock to invest in and technical analysis to determine the timing of the investment. Stock market prediction is a combination of several fields of study including statistics, operations research, computer science, economics, and finance. Traditionally, time series forecasting was used to predict the stock market that included classic regression methods that involved smoothing, moving average and autoregressive techniques [5]. More recently, multiple computational methods such as Bayesian Networks, Logistic Regression, Multiple Linear Regression (MLR), Support Vector Machines (SVM), Artificial Neural Networks (ANN), Genetic Algorithms (GA), and Case-Based Reasoning (CR) have been applied for forecasting the trends and movement of the stock market.