Course material & Lecture notes

Introduction to supervised learning, including problem formulation, datasets and labels, loss functions, and empirical risk minimization. Examples such as linear regression, with discussion of bias-variance tradeoff, overfitting, and ensemble methods (bagging and boosting).

Regularization in supervised learning, focusing on bias-variance tradeoff, overfitting and generalization. L1 and L2 regularization, ridge and lasso regression, and their effect on model complexity and stability.

Unsupervised learning and dimensionality reduction. PCA and eigenvalue interpretation, clustering methods such as k-means, and the curse of dimensionality. Overview of nonlinear techniques including t-SNE and UMAP.

Introduction to computer vision, covering image representation, color spaces, basic image processing, and classical vision tasks. Overview of modern deep learning-based vision pipelines, models, and training approaches.

Introduction to supervised learning, including problem formulation, datasets and labels, loss functions, and empirical risk minimization. Examples such as linear regression, with discussion of bias-variance tradeoff, overfitting, and ensemble methods (bagging and boosting).

Regularization in supervised learning, focusing on bias-variance tradeoff, overfitting and generalization. L1 and L2 regularization, ridge and lasso regression, and their effect on model complexity and stability.

Unsupervised learning and dimensionality reduction. PCA and eigenvalue interpretation, clustering methods such as k-means, and the curse of dimensionality. Overview of nonlinear techniques including t-SNE and UMAP.