Lectures for INFO8004 - Advanced Machine Learning, ULiège, Spring 2021.
- Instructors: Pierre Geurts ([email protected]), Gilles Louppe ([email protected]), Louis Wehenkel ([email protected])
- When: Spring 2021, Thursday 9:00 AM
- Classroom: Lifesize
Date | Topic |
---|---|
February 4 | Course syllabus [slides] Lecture 1: Gaussian and neural processes (Gilles Louppe) - Paper: "Conditional neural processes", Garnelo et al, 2018 [link] - Talk [slides] |
February 11 | Lecture 2: Statistical learning theory (Louis Wehenkel) - Tutorial & discussion: "Statistical Learning Theory - a Hitchhiker's Guide", John Shawe-Taylor and Omar Rivasplata, NeurIPS 2018 [video, slides]. - Additional resources: "An introduction to Statistical Learning Theory", Louis Wehenkel [slides]; "Statistical Learning Theory: A Primer", Louis Wehenkel [notes] |
February 18 | Lecture 3: Conformal prediction (Pierre Geurts) - Paper 1: "A tutorial on conformal prediction" , Shafer and Vovk, 2008 [link] - Paper 2: "Inductive conformal prediction: theory and application to neural networks", Papadopoulos, 2008 [link] - Paper 3: "Conformalized quantile regression", Romano et al, 2019 [link] - Talk [slides] |
February 25 | Lecture 4: Geometric Deep Learning - Tutorial: "Geometric Deep Learning", Michael Bronstein, MLSS 2020 [video] [slides] - Paper: "Geometric deep learning: going beyond Euclidean data", Bronstein et al, 2016 [link] |
March 4 | Lecture 5: Simulation-based inference (Gilles Louppe) - Paper: "The Frontier of Simulation-based Inference", Cranmer, Brehmer and Louppe, 2019 [link] - Talk [slides] |
March 11 | Lecture 6: Causality (Louis Wehenkel) - Paper: "Causality. Chapter 1: Introduction to Probabilities, Graphs, and Causal Models", Judea Pearl, 2009 [link] - Talk [slides] |
March 18 | Lecture 7: Upside-Down Reinforcement Learning (Matthia Sabatelli) - Paper 1: "Reinforcement Learning Upside Down", Juergen Schmidhuber, 2019 [link] - Paper 2: "Training Agents using Upside-Down Reinforcement Learning", Rupesh Kumar Srivastava et al, 2019 [link] - Talk [slides] |
March 25 | Lecture 8: Normalizing Flows (Antoine Wehenkel) - Paper: TBD - Talk [slides] |
April 1 | Lecture 9: Responsible AI (Adeline Decuyper, Accenture) |
April 15 | Student presentations 1 |
April 22 | Student presentations 2 |
April 29 | Student presentations 3 |
May 6 | Student presentations 4 |
Deliverables:
- a 30-minute lecture on the paper and its necessary background. This lecture will be presented to the class on April 15, April 22, April 29 or May 6.
- a summary report (4-6 pages) summarizing the problem, the contribution of the paper and a critical discussion. Use the
template-report.tex
template (although feel free to adjust if necessary).
Slides and reports should be submitted on the Montefiore submission platform. This assignment will count for 40% of the final grade.
The paper-group assignments are available here. If your name does not appear in the list, please contact us immediately, as otherwise you will be marked as Absent for the assignment.
TBD.