A Comparative Study of Crude Oil Futures Price Prediction Using Various Deep Learning Techniques
摘要
Crude oil price forecasting is a difficult task. Because international crude oil prices are nonlinear and change over time, As the backbone and genesis of countless businesses, oil plays a critical role in the global economy. It is an example of an key source of energy that serves as an essential raw material in many manufacturing operations and transportation. Oil prices are subject to severe volatility and swings. It is the most active and actively traded commodity on world markets. Much research has recently arisen to examine the topic of predicting oil prices and seeking the best outcomes. Despite these efforts, there were insufficient research to serve as a reference addressing all facets of the topic. In this paper we have taken crude oil future price from 2008 till 2023 and attempted to predict the Crude future prices by using Deep Learning techniques using “Naïve Model”, various models of “Keras”, “LSTM”, “N-Beats” and “Ensemble” model. It is a attempt to predict Crude oil Futures price as a Univariate analysis. Study has also compared between these models to consider which is a better prediction model among the rest.