.
Name: Ms. S.Lavanya
Ms.Register
Poorani.S
Number:192221116
Project
Guided by Approval
Dr. Mary Number G
Valantina. :SSE/26/21/116-04
Guided by- Dr. S.Jothi Shri
Crude Oil Price Prediction Using Random Forest Compared With Artificial Neural Network Quantitative
Algorithm
INTRODUCTION
Crude oil price prediction is crucial for various industries and economies. Machine learning algorithms offer powerful tools to analyze
historical data and forecast future prices. The aim of the study is to compare the effectiveness of Random Forest and Ann-Q machine
learning algorithms in predicting crude oil prices.
Recent advancements in machine learning have opened up new possibilities for predicting crude oil prices. Techniques like Random
Forest and Ann-Q offer more nuanced approaches to analyzing and forecasting price trends.
The volatility of crude oil prices is a central challenge in the energy market, impacting industries and economies worldwide. Predicting
these fluctuations accurately is crucial for decision-making and risk management.
Machine learning techniques, such as Random Forest and Artificial Neural Networks (ANN), offer powerful tools for forecasting crude
oil prices based on historical data and market indicators Fig.1:Crude oil price prediction using
Random Forest provides a clear measure of feature importance, offering insights into the factors driving crude oil prices. machine learning
MATERIALS AND METHODS
Random forest
Data collection Feature extraction
Collect historical data on Collect historical crude
crude oil prices along oil prices at different
with relevant features like time intervals. Visualisation
geopolitical events, Extract supply and Implementation prediction and
Artificial neural
economic indicators, and demand data such as of machine feature
network
supply-demand dynamics. production levels and learning importance
Preprocess the data by inventories. Include algorithms
handling missing values geopolitical events like
and scaling the features to conflicts or sanctions
a uniform range impacting oil markets. Accuracy value
Fig.2:Crude oil price prediction using machine learning predicted
RESULTS
Group Statistics
Accuracy Rate
Group N Mean Std. Deviation Std. Error
Mean
Random Forest 10 95.700 .48305 1.527
Accuracy
Artificial neural 10 94.500 .8498 .2687
1 2 3 4 5 6 7 8 9 10 11
network
Fig.3:A slice of accuracy comparison Fig.4:In the present work Random forest algorithm is compared
between Random Forest and Multilayer with Ann-q, and it depicts that the proposed algorithm gives Tab.1:Group Statistics comparison of RF and ANN
Perceptron using pie Chart more accuracy when compared with the existing algorithm.
DISCUSSION AND CONCLUSION
Based on T-test Statistical analysis, the significance value of p=0.001 (independent sample T - test p<0.05) is obtained and shows that there is a statistical significant
difference between the group 1 and group 2.Overall , the accuracy of the Random forest is 95.70% and it is better than the other algorithms artificial neural network
94.50%.
Random Forest consistently outperforms ANN in terms of accuracy. The ensemble nature of Random Forest allows it to capture complex patterns in the data more
effectively than ANN, leading to more accurate predictions. Random Forest provides a measure of feature importance, offering insights into the key drivers of crude oil
prices. This interpretability is valuable for stakeholders seeking to understand the factors influencing oil markets
Random Forest requires fewer computational resources and less training time compared to ANN, making it more practical for large-scale data analysis and deployment in
resource-constrained environments. Random Forest emerges as a superior choice for crude oil price prediction compared to ANN, offering higher accuracy, interpretability,
robustness, efficiency, and the ability to handle non-linear relationships.
BIBLIOGRAPHY
He, Xin James (2018). Crude Oil Price Forecasting: Time Series vs. SVR Models.Journal of International Technology and Information Management, 27 (2),25-43
Sircar, Anirbid, et al. "Application of machine learning and artificial intelligence in oil and gas industry." Petroleum Research 6.4 (2021): 379-391.
Escribano Á, Wang D. Mixed random forest, cointegration, and forecasting gasoline prices. International Journal of Forecasting. 2021 Oct 1;37(4):1442-62.
Mai-Cao, Lan, and Hoa Truong-Khac. "A Comparative Study on Different Machine Learning Algorithms for Petroleum Production Forecasting.(2022).Demirer, Riza, et al.
"Risk aversion and the predictability of crude oil market volatility: A forecasting experiment with random forests”. 73.8 (2022) 1755-1767.
Dong, Hao, Yingrong Zheng, and Na Li. "Crude oil futures price prediction by composite machine learning model."(2023).
Deng, Shangkun, Yingke Zhu, Shuangyang Duan, Yiting Yu, Zhe Fu, Jiahe Liu, Xiaoxue Yang, and Zonghua Liu. "High-frequency forecasting of the crude oil futures price with
multiple timeframe predictions fusion." Expert Systems with Applications 217 (2023): 119580.
