Encoding and Scaling Features

ML models need numerical inputs. Learn to encode categories and scale numbers.

Encoding Categorical Features

Feature Scaling

Key Takeaways

• Label encoding for ordinal categories (small/medium/large)
• One-hot encoding for nominal categories (red/green/blue)
• StandardScaler: mean=0, std=1 (most common)
• MinMaxScaler: scales to 0-1 range
• Many algorithms require scaling (SVM, KNN, Neural Nets)
• Tree-based models don't need scaling