AI-ML — Học máy & Trí tuệ nhân tạo

Lĩnh vực này dạy gì?

Foundations: pipeline, linear/logistic regression, bias-variance & regularization.
Classical methods: KNN, decision tree, random forest, SVM, gradient boosting — vẫn dominate tabular data.
Unsupervised: K-means, DBSCAN, PCA/SVD, t-SNE/UMAP.
Deep learning: MLP training deep, CNN, RNN/LSTM, Transformer + attention.
NLP & multimodal: text vectorization, word embeddings, vector DB + RAG, CLIP.

Tier 1

Pipeline ML end-to-end, linear regression deep (closed-form, regularization), logistic regression (softmax, calibration), bias-variance & cross-validation.

▶ Trang chính

Tier 2

Classical Methods

KNN & decision tree, random forest & bagging, SVM (kernel trick), gradient boosting (XGBoost/LightGBM).

▶ Trang chính

Tier 3

Unsupervised Learning

K-means & hierarchical, DBSCAN, PCA & SVD, t-SNE & UMAP — clustering + dimensionality reduction.

▶ Trang chính

Tier 4

Deep Learning

Neural network training deep (optimizers, batchnorm), CNN (conv/pool, ResNet), RNN/LSTM, Transformer + attention.

▶ Trang chính

Tier 5

NLP & Multimodal Applications

Text vectorization, word embeddings, vector DB + RAG, CLIP multimodal — ứng dụng NLP & LLM hiện đại.

▶ Trang chính

Lĩnh vực này dạy gì?

5 tier 20 bài, học tuần tự hoặc nhảy tier