Predicting Oil Price Trends During Conflict With Hybrid Machine Learning Techniques
The ongoing conflict between Russia and Ukraine has introduced significant volatility into the global oil markets, highlighting the need for robust forecasting models to understand and anticipate price fluctuations during such geopolitical events. This study presents a comprehensive hybrid modeling...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-01-01
|
| Series: | Applied Computational Intelligence and Soft Computing |
| Online Access: | http://dx.doi.org/10.1155/acis/8867520 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850235633782489088 |
|---|---|
| author | Hicham Boussatta Marouane Chihab Mohamed Chiny Younes Chihab |
| author_facet | Hicham Boussatta Marouane Chihab Mohamed Chiny Younes Chihab |
| author_sort | Hicham Boussatta |
| collection | DOAJ |
| description | The ongoing conflict between Russia and Ukraine has introduced significant volatility into the global oil markets, highlighting the need for robust forecasting models to understand and anticipate price fluctuations during such geopolitical events. This study presents a comprehensive hybrid modeling approach to predict oil prices in the context of Russia and Ukraine across three distinct periods: before the war, during the war, and the immediate aftermath of the conflict. Using advanced machine learning techniques, we developed a hybrid system combining Random Forest, ElasticNet, K-Nearest Neighbors, Gradient Boosting, and Support Vector Regressor models. These models were integrated through a Voting Regressor to enhance prediction accuracy. Our analysis revealed varying levels of predictive performance across the different periods. Prior to the war, the models showed strong predictive capabilities, evidenced by low mean-squared error (MSE) values and high coefficients of determination (R2). However, during the war, the models struggled to capture extreme volatility, resulting in significantly increased MSE and negative R2 values. Predictions for the immediate aftermath of the conflict demonstrated improvements, with a reduction in MSE and positive R2 values, indicating a return to relatively more stable market conditions. Notably, combining data from both before the war and war periods could further improve predictive accuracy, as it would reduce the impact of the conflict’s volatility on model performance. These results emphasize the challenges of forecasting oil prices in periods of geopolitical instability and underscore the importance of hybrid modeling approaches to adapt to rapidly changing market dynamics. |
| format | Article |
| id | doaj-art-3a83a50758024aba835cb77c6ac52e53 |
| institution | OA Journals |
| issn | 1687-9732 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Applied Computational Intelligence and Soft Computing |
| spelling | doaj-art-3a83a50758024aba835cb77c6ac52e532025-08-20T02:02:12ZengWileyApplied Computational Intelligence and Soft Computing1687-97322025-01-01202510.1155/acis/8867520Predicting Oil Price Trends During Conflict With Hybrid Machine Learning TechniquesHicham Boussatta0Marouane Chihab1Mohamed Chiny2Younes Chihab3Computer Science Research LaboratoryComputer Science Research LaboratoryComputer Science Research LaboratoryComputer Science Research LaboratoryThe ongoing conflict between Russia and Ukraine has introduced significant volatility into the global oil markets, highlighting the need for robust forecasting models to understand and anticipate price fluctuations during such geopolitical events. This study presents a comprehensive hybrid modeling approach to predict oil prices in the context of Russia and Ukraine across three distinct periods: before the war, during the war, and the immediate aftermath of the conflict. Using advanced machine learning techniques, we developed a hybrid system combining Random Forest, ElasticNet, K-Nearest Neighbors, Gradient Boosting, and Support Vector Regressor models. These models were integrated through a Voting Regressor to enhance prediction accuracy. Our analysis revealed varying levels of predictive performance across the different periods. Prior to the war, the models showed strong predictive capabilities, evidenced by low mean-squared error (MSE) values and high coefficients of determination (R2). However, during the war, the models struggled to capture extreme volatility, resulting in significantly increased MSE and negative R2 values. Predictions for the immediate aftermath of the conflict demonstrated improvements, with a reduction in MSE and positive R2 values, indicating a return to relatively more stable market conditions. Notably, combining data from both before the war and war periods could further improve predictive accuracy, as it would reduce the impact of the conflict’s volatility on model performance. These results emphasize the challenges of forecasting oil prices in periods of geopolitical instability and underscore the importance of hybrid modeling approaches to adapt to rapidly changing market dynamics.http://dx.doi.org/10.1155/acis/8867520 |
| spellingShingle | Hicham Boussatta Marouane Chihab Mohamed Chiny Younes Chihab Predicting Oil Price Trends During Conflict With Hybrid Machine Learning Techniques Applied Computational Intelligence and Soft Computing |
| title | Predicting Oil Price Trends During Conflict With Hybrid Machine Learning Techniques |
| title_full | Predicting Oil Price Trends During Conflict With Hybrid Machine Learning Techniques |
| title_fullStr | Predicting Oil Price Trends During Conflict With Hybrid Machine Learning Techniques |
| title_full_unstemmed | Predicting Oil Price Trends During Conflict With Hybrid Machine Learning Techniques |
| title_short | Predicting Oil Price Trends During Conflict With Hybrid Machine Learning Techniques |
| title_sort | predicting oil price trends during conflict with hybrid machine learning techniques |
| url | http://dx.doi.org/10.1155/acis/8867520 |
| work_keys_str_mv | AT hichamboussatta predictingoilpricetrendsduringconflictwithhybridmachinelearningtechniques AT marouanechihab predictingoilpricetrendsduringconflictwithhybridmachinelearningtechniques AT mohamedchiny predictingoilpricetrendsduringconflictwithhybridmachinelearningtechniques AT youneschihab predictingoilpricetrendsduringconflictwithhybridmachinelearningtechniques |