Forecasting stock index movement using Machine learning techniques
No Thumbnail Available
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
UMT,Lahore
Abstract
Due to economic conditions of every country stock market is gaining more and more importance day by day. As the stock prices are being considered the most important factor while making investment so prediction of stock price movement becomes necessary before investment. Stock market is very complex in nature and its Prediction has been remain an area of interest for many researchers. Change in stock market due to automatic buying and selling across the globe make it more daring to predict the market.
Predicting the direction of next closing price of Karachi stock exchange index is the aim of this study. Two supervised machine learning methods have been employed in the study for the prediction and also for comparison among these methods namely Random Forest (RF) and Support Vector Machine (SVM). Checking the best performance among the two algorithms is carried out by the process of cross-validation. In related literature there are many studies that have applied different machine learning techniques at different stock markets. According to best of my knowledge, comparison of Random Forest and Support Vector Machine in predicting the direction of Karachi stock exchange 100 index has not been examined before. Technical indicators computed from the closing price, opening price, high price and low price are used as input variables for this study. 70% of data has been used to train the models and generate predicted labels as output. Then predicted labels reveal the direction of the closing price of the next day. This will create hit rate for both algorithms on which basis the best model can be selected. Results for this study revealed that among both algorithms random forest outperforms the support vector machine in regards of the performance. Support vector machine has better predicted the direction for stock market with the 79.76% hit ratio.