Module 05
Classical Machine Learning
Tree ensembles still win most tabular problems in industry. Learn this properly before deep learning, not after.
22 topics ~104 h estimated learning time
Sign in to keep your progress.
Ready to start
- 5.1 Framing an ML problem
Foundation · 3 h
- 5.2 Linear regression from scratch
Foundation · 6 h
- 5.3 Logistic regression and classification
Foundation · 6 h
- 5.4 Regularization
Core · 4 h
- 5.5 Bias, variance and model capacity
Foundation · 4 h
- 5.6 Cross-validation and honest evaluation
Core · 5 h
- 5.7 Metrics and calibration
Core · 5 h
- 5.8 k-nearest neighbours and distance methods
Foundation · 3 h
- 5.9 Naive Bayes
Foundation · 3 h
- 5.10 Support vector machines and kernels
Core · 5 h
- 5.11 Decision trees
Foundation · 4 h
- 5.12 Bagging and random forests
Core · 5 h
- 5.13 Gradient boosting: XGBoost, LightGBM, CatBoost
Core · 7 h
- 5.14 Clustering
Core · 5 h
- 5.15 Dimensionality reduction
Core · 5 h
- 5.16 Anomaly detection
Core · 4 h
- 5.17 Imbalanced data
Core · 4 h
- 5.18 Hyperparameter optimization
Core · 4 h
- 5.19 scikit-learn pipelines end to end
Core · 5 h
- 5.20 Model interpretability
Core · 5 h
- 5.21 Time-series forecasting
Core · 7 h
- 5.22 Recommender systems
Core · 5 h