Chih-Jen Lin

National Taiwan University
  1. Introduction: why optimization and machine learning are related
  2. Main optimization techniques for kernel methods. Main optimization techniques for linear classification. Other optimization techniques used in kernel or linear classification
  3. Lessons learned from working on both areas (optimization and machine learning)
  4. Optimization techniques for multi-core/distributed training
  5. Slides
  1. Representation Learning & Fundamentals of Generative Adversarial Networks (GANs)
  2. GAN Training - Optimization issues
  3. Taxonomy of GANs & Applications
  4. GANs for Discrete data such as Text
  5. Slides

Laura Palagi

Sapienza University of Rome
  1. Feedforward Neural Networks: optimization issues
  2. Decomposition w.r.t. number of samples: first order incremental/stochastic methods
  3. Decomposition w.r.t. number of samples: second order incremental/stochastic methods
  4. Decomposition w.r.t. the layers: extreme learning and beyond
  5. Slides

Massimiliano Pontil

Istituto Italiano di Tecnologia & University College London
  1. Introduction to statistical learning
  2. Multitask learning I
  3. Multitask learning II
  4. Hyperparameter optimization
  5. Slides