DATS0001 Foundations of Data Science

Materials for DATS0001 Foundations of Data Science, ULiège, Fall 2023.

• Instructor: Gilles Louppe
• When: Monday 1:30 PM
• Classroom: B28/1.21
• Discord: https://discord.gg/qpvh4ueEFQ

Agenda

Date	Topic
September 18	No class
September 25	Course syllabus Lecture 1: Introduction nb01: Build, compute, critique, repeat [notebook] Reading: Blei, Build, Compute, Critique, Repeat, 2014 [Section 1] Reading: Box, Science and Statistics, 1976
October 2	Lecture 2: Data nb02a: Tables [notebook] nb02b: JAX [notebook] nb02c: Data wrangling [notebook] Reading: Harris et al, Array programming with NumPy, 2020
October 9	Lecture 3: Visualization nb03a: Plots [notebook] nb03b: Data visualization principles [notebook] nb03c: High-dimensional data [notebook] Reading: Rougier et al, Ten Simple Rules for Better Figures, 2014 Reading: Rougier, Scientific Visualization: Python+Matplotlib, 2022
October 16	No class
October 23	Lecture 4: Bayesian modeling nb04: Latent variable models [notebook, sidenotes (LVMs), sidenotes (Probabilistic PCA)] Reading: Gelman et al, Bayesian workflow, 2020 [Sections 1 and 2] Reading: Blei, Build, Compute, Critique, Repeat, 2014 [Sections 2 and 3]
October 30	No class
November 6	Lecture 5: Markov Chain Monte Carlo nb05: Markov Chain Monte Carlo [notebook] [sidenotes] Reading: Gelman et al, Bayesian Data Analysis, 3rd, 2021 [Chapter 11]
November 13	Lecture 6: Expectation-Minimization nb06: Expectation-Maximization [notebook] [sidenotes] Reading: Dempster et al, Maximum Likelihood from Incomplete Data via EM, 1977
November 20	Lecture 7: Variational inference nb07: ADVI [notebook] [sidenotes] Reading: Kucukelbir et al, Automatic Differentiation Variational Inference, 2016
November 27	Lecture 8: Model criticism nb08a: Model checking [notebook] nb08b: Model comparison [notebook] Reading: Gelman et al, Bayesian Data Analysis, 3rd, 2021 [Chapters 6 and 7]
December 4	Lecture 9: Wrap-up case study nb09: Space Shuttle Challenger disaster [notebook] Reading: Cam Davidson-Pilon, Bayesian Methods for Hackers, 2015 [Chapter 2]

Homeworks

• Homework 1: Exploration of solar power data and weather data (due by November 6)
• Homework 2: Modeling photovoltaic power production (due by December 1)
• Homework 3: Improving and comparing forecasting models (due by December 15)
• Exam-at-home: TBD

Homeworks must be submitted on Github classroom. Follow the links sent by email to register to each homework.