classical-ml
Unsupervised learning
Clustering, dimensionality reduction, density estimation — finding structure in unlabelled data.
Уровни глубины
L0Intro~2ч
Knows what clustering and PCA are; understands why unlabelled data is useful.
L1Basics~15ч
Applies k-means, DBSCAN, PCA, t-SNE; evaluates clustering with silhouette score.
L2Working~25ч
Uses GMMs, ICA, autoencoders; selects appropriate algorithms for structure discovery.
L3Advanced~40ч
Derives EM algorithm; applies manifold learning methods; works with self-supervised pretraining.
L4Research~80ч
Contributes to representation learning, self-supervised methods, or generative modelling research.
Ресурсы
L1 — Basics