Stock Direction Forecasting Techniques: An Empirical Study Combining Machine Learning System with Market Indicators in the Indian Context

Abstract

Stock price movement prediction has been one of the most challenging issues in finance since the time immemorial. Many researchers in past have carried out extensive studies with the intention of investigating the approaches that uncover the hidden information in stock market data. As a result of which, Artificial Intelligence and data mining techniques have come to the forefront because of their ability to map nonlinear data. The study encapsulates market indicators with AI techniques to generate useful extracts to improve decisions under conditions of uncertainty. Three approaches (fundamental model, technical indicators model and hybrid model) have been tested using the standalone and integrated machine learning algorithms viz. SVM, ANN, GA-SVM, and GA-ANN and the results of all the three approaches have been compared in the four above mentioned methods. The core objective of this paper is to identify an approach from the above mentioned algorithms that best predicts the Indian stocks price movement. It is observed from the results that the use of GA significantly increases the accuracy of ANN and that the use of technical analysis with SVM and ANN is well suited for Indian stocks and can help investors and traders maximize their quarterly profits.

Performance Forecasting of Share Market using Machine Learning Techniques: A Review

“…It can also be considered as an indication of realizing share growth potential [1], [6], [17]. [6], [12][13], [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37]. The main idea behind using these indicators is to evaluate stock price movements based on historical price patterns and volumes.…”

“…It is a supervised learning algorithm and initially it was designed to solve pattern recognition problems, but it has rendered to solve non -linear regression problems as well [23], [52]. Table 5 shows the description of the SVM technique as selected by most of the authors for share forecasting.…”

Performance Forecasting of Share Market using Machine Learning Techniques: A Review

“…Next [6], [12][13], [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37]. The main idea behind using these indicators is to evaluate stock price movements based on historical price patterns and volumes.…”

Predicting stock market trends using random forests: A sample of the Zagreb stock exchange

“…Choudhry and Garg [7] proposed a hybrid genetic-algorithm SVM-model for predicting the direction of stock price movement in the Indian stock market. Chandwani and Singh Saluja [8] used a combination of both fundamental and technical indicators and applied genetic-algorithm SVM-and ANNhybrid models to predict the direction of 25 companies listed on the Bombay Stock Exchange. Patel et al [9] discretised technical indicators by representing them as trend deterministic data and then trained predictive models using random forests, ANNs, naïve Bayes and SVM models on CNX Nifty, S&P BSE Sense indices and…”