KEV0143/Comparative-analysis-of-hourly-load-forecasting-using-PatchTST-TFT-NHiTS-and-CatBoost
Benchmark comparing four time-series forecasting models (three deep learning, one gradient boosting) for 24-hour energy load prediction.
Velocity · 7d
+4.4
★ / day
Trend
→steady
star history
This repository evaluates state-of-the-art deep learning architectures against traditional machine learning for hourly load forecasting. It implements PatchTST, Temporal Fusion Transformer (TFT), and N-HiTS alongside CatBoost, providing data preprocessing, model training, validation, and result visualization pipelines. The benchmark supports decision-making in energy markets by comparing model accuracy across standardized evaluation metrics.