A robust multi-location evaluation of a machine learning framework for wind power forecasting
Usman Ali, Muhammad Sufyan, Shahzad Ali, Mudassar Ahmad, Sajawal ur Rehman Khan, Naeem Raza, Jabeen Sultana, Muhammad Asif Habib · 2026
Read A robust multi-location evaluation of a machine learning framework for wind power forecasting by Usman Ali, Muhammad Sufyan, Shahzad Ali, Mudassar Ahmad, Sajawal ur Rehman Khan, Naeem Raza, Jabeen Sultana, Muhammad Asif Habib (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0344971.txt.
Description
Usman Ali and colleagues benchmarked machine learning methods for forecasting wind power across four sites in different countries. Grid operators need these predictions to balance supply. The team cleaned outliers from each dataset, then compared gradient boosting, random forest, and support vector machines with three kernel styles, judging them on accuracy and average error.
Gradient boosting aced every site. Random forest matched it in three places but fell apart at the fourth, with errors ballooning. Among support vector variants, the simplest linear version quietly excelled everywhere, while fancier kernels lagged. The authors recommend the two winners as dependable choices across varied wind climates.
Catalog record
| CREATOR | Usman Ali, Muhammad Sufyan, Shahzad Ali, Mudassar Ahmad, Sajawal ur Rehman Khan, Naeem Raza, Jabeen Sultana, Muhammad Asif Habib |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 100237 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |