Fading Coder

One Final Commit for the Last Sprint

Time Series Feature Engineering for Electricity Demand Forecasting

Core Feature Categories in Time Series Prediction Effective time series forecasting relies heavily on feature engineering. The following categories represent essential feature types that capture different temporal patterns in electricity demand data. 1. Temporal Decomposition Features Date-based fea...

Introduction to LightGBM and Feature Engineering for Electricity Demand Prediction

Introduction to LightGBM and Feature Engineering for Electricity Demand Prediction
1. Learning Objectives Plot bar charts and line charts using the dataset. Construct historical shift features and window statistical features from time series data. Train and predict using the LightGBM model. 2. GBDT and LightGBM GBDT (Gradient Boosting Decision Tree) is a long-standing model in mac...