Notes, and implementations on Supervised Learning. Covers the foundations and modern practice of supervised learning, from classical algorithms to deep learning, object detection, and Transformers.
Supervised Learning Foundations
- Formulation of the learning process
- Classification and regression frameworks
- Experiment design: dataset splits, metrics, augmentation
- Model evaluation and comparison
Classical Algorithms
- LDA (Linear Discriminant Analysis)
- Decision Trees
- k-Nearest Neighbors (k-NN)
- Support Vector Machines (SVM)
- Neural Networks
Ensemble Methods
- Bagging and Boosting
- Random Tree Ensembles
- Stacking
Classical Computer Vision
- Local descriptors: SIFT and Bag of Words (BoW)
Deep Learning
- Convolutional Neural Networks (CNNs): training, famous architectures, transfer learning
- Recurrent and Recursive Networks (RNNs, LSTM, GRU)
- Transformers
- Self supervised learning
Modern Object Detection
- Two-stage: R-CNN, Fast R-CNN, Faster R-CNN
- One-stage: YOLO