Seonglae ChoModels like linear regression, SVMs, and neural networks are sensitive to scaling (not tree-based model)
After data splitting, the scaler should only be fit and transformed on the training data, while test data should only be transformed using the scaler fitted on training data.