Predictive analysis comparison of energy stock prices using machine learning models

• Main Author: Valentino, Egan
• Other Authors: Wong Jia Yiing, Patricia
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/167817
• Institution: Nanyang Technological University

Recurrent neural networks (RNN), Long Short-Term Memory (LSTM), and Auto Regressive Integrated Moving Average (ARIMA) are some of the machine learning models that have demonstrated potential in forecasting temporal sequences, including the prices of stocks or commodities, by analyzing past data and incorporating various factors like technical indicators. As the stock market is complex and unpredictable due to several unknown factors, even experienced analysts find it challenging to predict market trends. However, there is a growing interest in applying machine learning to predict stock prices, as a model that can accurately forecast future prices could significantly impact investment firms' and traders' profits. This report compares various machine learning models for predicting the stock prices of energy companies. With the recent emergence of the energy transition, the energy industry is gaining more traction and the market will only grow bigger over the years. To research this arising market, the study examines the performance of different models, including regression, random forests, and neural networks. All these analyses use historical data and other factors on energy stocks and evaluate the models' accuracy in predicting future stock prices. The final aim is to investigate which machine learning models through different analysis and literature review would result in the most accurate price prediction possible for different types of energy companies in different scenarios.