Smart grids and machine learning: Uncovering pitfalls and improving the robustness of models

2026-07-10

Emran Altamimi *

Hussein Aly

Abdulaziz Al‐Ali

Abdulla Al‐Ali

Qutaibah Malluhi

DOI：10.1016/j.egyr.2026.109483

Original

Original request for help

Save

Abstract

En 中文

• Identify key errors in smart grid ML, such as flawed preprocessing and evaluation. • Improper scaling, splitting, and sampling cause data leakage and inflated metrics. • Tuning hyperparameters on labeled data breaks the core unsupervised assumption. • Best-practice workflow to mitigate these pitfalls for real-world applicability.

Keywords:

Smart grids

Machine learning

Privacy

Data preprocessing

Unsupervised learning

Smart grids and machine learning: Uncovering pitfalls and improving the robustness of models

Abstract

Journal

Organization

Cited Papers

Citing Papers