Application of machine learning in stock index forecast

Stock price prediction is an extremely complex problem due to the numerous factors and events occurring around the world, and yet has always been a popular topic of research. In recent times, advancement in machine learning techniques has sparked new ideas and possibilities that utilize these advanced techniques for the task of stock price prediction. In this project, I will be using the Tesla, Inc. stock (NASDAQ: TSLA) as the subject of my research. Technical features of this stock were retrieved from widely available finance sites such as Yahoo Finance, and content features were obtained from tweets from selected accounts of the popular social media site, Twitter. Content features were pre-processed and transformed into sentiment scores and afterwards, the sentiment scores together with the technical features were fed into a machine learning model to predict the price movement of TSLA. The proposed machine learning model is a Long Short-Term Memory (LSTM) neural network, and it is evaluated against traditional/statistical models such as the Autoregressive Integrated Moving Average (ARIMA) model which accepts univariate data inputs, as well as the Vector Autoregression (VAR) model which accepts multivariate data inputs.Bachelor of Engineering (Computer Science