"Impact of Climate Variability on the Performance of Supervised Machine Learning Models in Renewable Energy Forecasting"

EasyChair Preprint 14448

Abstract

The growing reliance on renewable energy sources, such as solar and wind power, has introduced new challenges in accurately forecasting energy production due to the inherent variability of these resources. Climate variability, characterized by fluctuations in weather patterns and extreme events, directly affects the performance of supervised machine learning (ML) models used for renewable energy forecasting. This research investigates the impact of climate variability on the accuracy, robustness, and generalizability of supervised ML models in the context of renewable energy forecasting.

By analyzing historical climate data and energy production records across various geographic regions, this study aims to identify the key climate factors that influence model performance. The research employs a range of supervised ML models, including neural networks, support vector machines, and ensemble methods, to forecast energy output under different climatic conditions. The study also explores the resilience of these models to climate-induced anomalies and evaluates their adaptability in scenarios of increasing climate variability.

Ultimately, this research provides a comprehensive understanding of the interplay between climate variability and ML model performance, offering valuable insights for the development of more resilient forecasting systems in the face of changing climatic conditions.

Keyphrases: Climate variability, Grid Management, Supervised Machine Learning, Support Vector Machines, climate-aware features, ensemble methods, extreme weather events, model performance, neural networks, renewable energy forecasting, sustainable energy systems

@booklet{EasyChair:14448,
  author    = {Adeoye Ibrahim},
  title     = {"Impact of Climate Variability on the Performance of Supervised Machine Learning Models in Renewable Energy Forecasting"},
  howpublished = {EasyChair Preprint 14448},
  year      = {EasyChair, 2024}}