Advanced Data Analytics, HEC Lausanne, Fall Semster 2025

Data Scientist (n.): Person who is better at statistics than any software engineer and better at software engineering than any statistician.

• This is a 14-week Master’s level course on machine learning, offered at HEC Lausanne during the Fall Semester 2025 (Monday, September 15 - Monday, December 15, 2025).
• The course format consists of three 45-minute lecture sessions, followed by a 45-minute hands-on session at the end of each lecture.

Meeting time and location

• Meeting time: Mondays, 10:15 - 12:00; 16:30 - 18:00
• Where: Internef 126 (10:15 - 12:00); Anthropole 3185 (16:30 - 18:00)

TA sessions:

• A weekly hands-on session is held on Mondays from 17:15 to 18:00, led by the teaching assistant, Maria Pia Lombardo.
• Individual TA sessions can be arranged via email and take place on Fridays upon request.

Class enrollment on the Nuvolos Cloud

• All lecture materials (slides, codes, and further readings) will be distributed via the Nuvolos Cloud.
• To enroll in this class, please click on this enrollment key, and follow the steps.

Video to get started with Nuvolos

• First steps on Nuvolos (Video available on Nuvolos platform)

Course Objectives

• The objective of the course is to gain practical familiarity with current computer-aided data analysis and machine-learning approaches.

• We will study various machine learning methods in the context of:
  • supervised machine learning (classification and regression).
  • unsupervised machine learning.
  • reinforcement learning (if time permits).
• The lectures will be interactive, in a workshop-like style, that is, a mix of theory and actively playing with code examples (delivered in Python and deployed on a cloud computing infrastructure).

Prerequisites

• Basic econometrics.
• Basic programming in Python (see this link for a thorough introduction).
• Basic calculus and probability (Mathematics for Machine learning provides a good overview of skills participants are required to be fluent in).

Approximate schedule

Witness the incredible transformation of a student throughout the Machine Learning course, from a blank slate to a pioneer in AI!

Week 1, Monday, September 15th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Get up to speed – week 1.
  • Topics:
    • Introduction to Machine Learning
    • A brief setup session on Nuvolos

Week 2, Monday, September 22th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• No Class: Swiss Federal Fast, Public Holiday

Week 3, Monday, September 29th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Get up to speed – week 2.
  • Topics:
    • A crash course on Programming in Python
    • Introduction to Jupyter Notebooks
    • A brief introduction to GitHub (quick tour; self-study)
    • Finger exercises (at the end of the lecture; self-study)
    • Introduction to Numpy (quick tour; self-study)
    • Introduction to Scipy (quick tour; self-study)
• Distribution of Exercise sheet 1

Week 4, Monday, October 6th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 4
  • Topics:
    • Supervised Learning - the general idea
    • Linear Regression (with multiple variables)
    • Gradient Descent
    • Polynomial Regression
    • Tuning Model Complexity
    • Stock Market Prediction (if time permits)
    • Introduction to Pandas (quick tour; self-study)
    • Discussion of the previous exercises; Finger exercises (with TA)
• Distribution of Exercise sheet 2

Week 5, Monday, October 13th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Intermezzo: a dedicated hands-on session on programming with Generative AI (LLMs) and autonomous agents, taught by Anna Smirnova.
• Lecture slides, week 5

Week 6, Monday, October 20th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 6
  • Topics:
    • Supervised Learning: Classification
    • k-Nearest-Neighbours
    • How to evaluate Classifiers
    • Naive Bayes
    • Decision Trees
    • Combining Models (Boosting, Bagging – if time permits)
    • Introduction to matplotlib (if time permits)
    • Discussion of the previous exercises; Finger exercises (with TA)
• Distribution of Exercise sheet 3

Week 7, Monday, October 27th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 7
  • Topics:
    • Deep learning basics
    • Multi-layer perceptron
    • Feed-forward networks
    • Network training - SGD
    • Error back-propagation
    • Some notes on overfitting
    • Introduction to Tensorflow, applied to supervised machine learning problems
    • Discussion of the previous exercises (with TA)
• Distribution of Exercise sheet 4

Week 8, Monday, November 3rd, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 8
  • Topics:
    • Warm-up with Tensorflow: Approximate analytical functions
    • A comprehensive Tour to Tensorflow
    • Introduction to PyTorch, applied to supervised machine learning problems
    • Introduction to Deep Learning for Sequence Modeling (Recurrent Neural Networks, LSTMs, and a first glimpse into self-attention)
    • LSTMs applied to Ozone time series data, and stock prices (I)
    • Discussion of the previous exercises; Finger exercises (with TA)
• Distribution of Exercise sheet 5

Week 9, Monday, November 10th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Recap week – the most important concepts of the first 8 weeks will be reviewed, and open questions discussed so that no one is left behind
  • Topics:
    • Recap of Supervised Machine Learning: Regression
    • Recap of Supervised Learning: Classification
    • Recap of Stochastic Gradient Descent
    • Feed-forward networks
    • Sequence Modeling with Neural Networks (RNNs, LSTMs)
    • Applications in NLP (at the end of the lecture; self-study)
    • Discussion of the previous exercises (with TA)

Week 10, Monday, November 17th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 10
  • Topics:
    • Gaussian Process regression
    • Kernels with noise
    • Option-pricing examples (if time permits)
    • GP classification (self-study)
    • Finger exercises part 1 (at the end of the lecture (with TA))
    • Finger exercises part 2 (at the end of the lecture (with TA))
• Distribution of Exercise sheet 6

Week 11, Monday, November 24th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 11
  • Topics:
    • The curse of Dimensionality
    • Bayesian active learning
    • Active Subspaces and Gaussian Process Regression
    • Principal Component Analysis
    • Discussion of the previous exercises; Finger exercises (with TA)
• Distribution of Exercise sheet 7

Week 12, Monday, December 1st, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 12
  • Topics:
    • k-Means
    • Gaussian Mixture Models
    • Expectation Maximization
    • Hierarchical Clustering
    • Density-based Clustering
    • Discussion of the previous exercises; Finger exercises (with TA)
• Distribution of Exercise sheet 8

Week 13, Monday, December 8th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 13
  • Topics:
    • Introduction to Reinforcement learning (RL)
    • A notebook on RL basics
    • A notebook on Q learning
    • RL: Optimal portfolio choice application
    • Discussion of the previous exercises (with TA)

Week 14, Monday, December 15th, 2025 (10:15 - 12:00; 16:30 - 18:00)

• Lecture slides, week 14
  • Topics:
    • An Application of the tools studied so far: Deep Uncertainty Quantification for Integrated Assessment Models (Supervised, Unsupervised, and Reinforcement Learning)
    • Finger exercises (at the end of the lecture; self-study)
    • Wrap-up of the lecture
    • Voluntary project presentations.
    • 2h Consulting for the captstone project (with TA)

Grading

• Every student has to provide a capstone project that illustrates what was learned.
• Each student individually has to propose a data science project and work on it over the course of the semester.
• The due date to submit the project is in the last week of the semester.
• The deliverables are:
  • i) a report of about 10 pages lengths.
  • ii) a GitHub repository with the related code and data.
  • iii) a video recording of a maximum of 10 minutes length that presents the project, the findings, etc.
• We will award the grades based on whether the captstone project demonstrates an understanding of the material. There will be no exams.

Lecturer

• Simon Scheidegger (University of Lausanne, Department of Economics)
  • Simon Scheidegger: simon.scheidegger@unil.ch

TAs and support

• Maria Pia Lombardo mariapia.lombardo@unil.ch
• Anna Smirnova: anna.smirnova@unil.ch
• Nuvolos Support: support@nuvolos.cloud

Google document for questions on the previous lecture:

• Google Doc

References

• An Introduction to Statistical Learning (James, Witten, Hastie, Tibshirani; Springer)
• Deep Learning (Goodfellow, Bengio, Courville; MIT Press)
• Pattern Recognition and Machine Learning (Bishop; Springer)

Auxiliary materials

• Git intro (Screencast available on Nuvolos)
• Terminal intro (Screencast available on Nuvolos)