0 of 4 done
The algorithm zoo — the non-deep-learning models that are still the right tool for most structured/tabular problems, often beating neural networks there while training in seconds and staying interpretable. Part of the ML Engineer track.
- Regression & Regularization — [Intermediate] — linear, polynomial, and logistic regression; ridge/lasso/elastic-net regularization
- Trees & Ensembles — [Intermediate] — decision trees, random forests, and gradient boosting (XGBoost/LightGBM) — the tabular-data workhorses
- SVM, KNN & Naive Bayes — [Intermediate] — three more classic supervised algorithms and when each fits
- Clustering & PCA — [Intermediate] — unsupervised learning: k-means/hierarchical clustering and dimensionality reduction (PCA)
Related
- Model Evaluation — how to judge these models
- Deep Learning — when to reach past classical ML
- Mathematics — the linear algebra/stats underneath
Jump to a subject
Software engineering
Infrastructure & security
Computer science
Hardware & engineering