Unlike most other machine learning regression systems, when using LightGBM, numeric predictor and target variables can be used as-is. You can normalize numeric predictors using min-max, z-score, or ...