Prediction of stock prices using hybrid machine learning models

Stock price prediction is one of the highly discussed, debated, and researched areas in the realm of investments. Investors are highly interested in stock price prediction strategies to increase their profitability in stock trading and reap dividends by investing in potential stocks. However, the level of accuracy of stock price predictions plays a key role in determining the outcome of the trading strategy of the investors. Thus, the accuracy of the stock price prediction method is of utmost significance for investors in the stock market. There is a multitude of prediction methods based on technical and fundamental analysis. However, with the ever-increasing data generation during transactions and high levels of market volatility, machine learning models have been tasked with predicting stock prices based on an array of information. Stock price prediction is a time series forecast problem that can be handled by a multitude of machine learning models with varying levels of prediction accuracy. Further, due to the uniqueness of the behavior of individual stock markets, different machine learning models should be deployed in different stock markets to generate the most accurate predictions. In consideration of the significance of stock price prediction and the major role assigned to machine learning models to be played in maximizing investor profitability, it is pertinent to explore the performance of different machine learning models in predicting stock prices in the Sri Lankan context. Apart from exploring the prediction accuracy of different machine learning models, it is beneficial to further explore the new concepts of deploying hybrid machine learning models for stock price predictions,