Syllabus

Introduction (ML applications)-topics in Linear system (linear regression)-Basics of Gradient Descent and its variants (logistic regression)-A detailed understanding of Projected Gradient (white-box adversarial attack)and Proximal Gradient (lasso)-Details of Conditional Gradient (recommendation system)-The Sub gradient approach (svm)-Mirror Descent and Metric Gradient methods-Acceleration (total variation denoising)-Smoothing (robust svm)-optimal transport for machine learning -Alternating (VAE)- Minimax (adversarial training)-Averaging (GANs)-Splitting (federated learning)-Extra gradient (maxentropy)-Stochastic Gradient (Boltzmann machine)-Variance Reduction (boosting)- Derivative- free (black-box adversarial attack)