A Comparative Analysis of Different Machine Learning Algorithms Used in Predicting Stock Market Prices

Abstract: Stock markets can be defined as a trading platform for the exchange of financial instruments such as debt, equity and derivatives. They work on the principle of price discovery, which is the act of studying the market supply and demand of a commodity and determining its proper price. Essentially when a person is investing in the stock markets, there are two kinds of situations that are prevalent. One is uncertainty and the other is risk. A person would hedge a bet only if the proposition is risky, however if things turn out to be uncertain, he would be averse to investing in the stock. By studying different machine learning algorithms, we analyse the best methods available in predicting the movement of stocks.

Keywords: Neural Networks, Genetic Algorithm, Regression, Decision Tree, SVM.