Improving Solar Radiation Prediction with Cheetah Optimizer and Random Forest
Author Information
Author(s): Al-Shourbaji Ibrahim, Kachare Pramod H., Jabbari Abdoh, Kirner Raimund, Puri Digambar, Mehanawi Mostafa, Alameen Abdalla
Primary Institution: Department of Electrical and Electronics Engineering, Jazan University, Jazan, Saudi Arabia
Hypothesis
Can the Cheetah Optimizer improve the accuracy of solar radiation predictions when combined with Random Forest?
Conclusion
The Cheetah Optimizer-Random Forest model significantly outperforms other machine learning techniques in predicting solar radiation.
Supporting Evidence
- The CO-RF model achieved a low MAE of 0.0365 and an R2 of 0.9251 on the first dataset.
- The CO-RF model outperformed Logistic Regression, Support Vector Machine, and standalone Random Forest.
- On the second dataset, the CO-RF model achieved an MAE of 0.0469 and an R2 of 0.9868.
Takeaway
This study shows a new way to predict how much sunlight will hit the ground using a special computer program that learns from data.
Methodology
The study used two publicly available solar radiation datasets and evaluated the Cheetah Optimizer-Random Forest model using metrics like Mean Absolute Error and R-squared.
Limitations
The study is limited to two datasets and may not generalize to all geographical areas.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website