HS2023: 42032/62032 Machine Learning and Data Mining

This course gives an introduction to the theory of machine learning. Application is in the form of a course project.
During the course, you will be able to:

- Formulate machine learning problems in terms of opimisation or probabilistic inference.
- Apply off-the-shelf algorithms to problems with scikit-learn
- Develop custom models and algorithms using python libraries such as pyTorch

Grading:
- Exam: 40%
- Project: 40%
- Assignments: 20%

1. Learning as optimisation
1.1 Objective functions
1.2 k-Nearest neighbours
1.3 Generalisation
1.4 Linear neural networks
1.5 Stochastic gradient descent
1.6 Multi-layer neural networks
1.7 Backpropagation
2. Learning as probabilistic inference
2.1 Probabilistic models
2.2 Discriminative models
2.3 Generative models
2.4 Expectation Maximisation
2.5 Monte-Carlo Inference
2.6 Variational inference
3. Sequence modelling
3.1. Auto-regresive models
3.1.1 MLP Auto-Regressors and Transformers
3.1.2 (Variable Order) Markov Models
3.2. Recursive models:
3.2.1. Recursive Neural Networks
3.2.2. Hidden Markov Models
4. Decision making
4.1. General statistical decision problems
4.2. Reinforcement learning

Deutsch

All rights reserved

Christos Dimitrakakis

christos.dimitrakakis@unine.ch

Fridays 13:00-14:00, B116, by appointment.

Unbegrenzt – wenn online geschaltet

Sie können diesem Kurs direkt beitreten.

Unbegrenzt

Anmeldename
Vorname
Nachname
E-Mail