Crude Oil Price Forecasting Model Using Machine Learning
Crude Oil Price Forecasting Model Using Machine Learning Tapas Peshin1 and
Nikolaos V. Sahinidis2 1Graduate Student, Department of Chemical Engineering,
Carnegie Mellon University, Pittsburgh, USA
[email protected] 2John E.
Swearingen Professor, Department of Chemical Engineering, Carnegie Mellon
University, Pittsburgh, USA
[email protected] ABSTRACT The impact of oil
price on the social, economic, political and many other aspects of human life is
substantial. Oil exemplifies a vital role in the world economy as the backbone and
origin of numerous industries. In global markets, it is the most active and heavily
traded commodity. Global oil prices have fallen sharply over the past few months,
leading to significant revenue shortfalls in many energy exporting nations. From
2010 until mid 2014, world oil prices had been fairly stable. WTI crude oil has
declined ~58% since the middle of June 2014. Brent crude oil has also declined
~69% from mid June. Recently many studies have emerged to discuss the problem
of predicting oil prices and seeking to access to the best outcomes. This paper
focuses on the use of a data driven approach to predict crude oil prices using
Automated Learning of Algebraic Models for Optimization (ALAMO) and
comparing its results with the tools and techniques used in the past. Keywords:
Crude Oil Price Prediction, ALAMO, Modeling, Surrogate Modeling (SUMO)
Toolbox, Neural Network Time Series INTRODUCTION Crude oil is the most
important