Feature Engineering & Selection — Machine Learning Roadmap
Often the single biggest lever for improving real-world model performance
Steps in Feature Engineering & Selection
- Feature Creation & Domain-Driven Features — advanced · Engineering new, more predictive features from raw data using domain knowledge
- Feature Selection Techniques — advanced · Filter, wrapper and embedded methods for choosing which features actually matter
- Handling Imbalanced Datasets — advanced · Oversampling, undersampling and SMOTE for datasets with rare positive classes
- Text Feature Engineering (Bag of Words, TF-IDF) — advanced · Turning raw text into numeric features a classical ML model can use
- Time-Based Feature Engineering — advanced · Extracting lag features, rolling statistics and seasonality from time series data
Part of
- Machine Learning roadmap — the full learning path