TIES4700 Deep Learning (5 cr)

Study level:
Advanced studies
Grading scale:
0-5
Language:
English, Finnish
Responsible organisation:
Faculty of Information Technology
Curriculum periods:
2024-2025, 2025-2026, 2026-2027, 2027-2028

Description

This course takes the student into the world of deep learning, focusing on the most common deep learning methods, such as various neural networks (CNN, RNN, Transformer, GAN). The course deepens students' understanding of the mathematical models of deep learning, optimization algorithms, computational requirements, and hyperparameter optimization. During the course, the textbook 'Dive into Deep Learning' (https://d2l.ai/) is used, which provides students with a thorough understanding of the key concepts and techniques of deep learning. The course combines theory and practice, giving students the ability to apply deep learning to complex data-driven problems.

Learning outcomes

After completing the course, the student is familiar with the most common deep learning methods (CNN, RNN, Transformer, GAN) and understands the mathematical models related to deep learning, optimization algorithms, computational requirements, and hyperparameter optimization. The student can assess when a model can be expected to generalize to unseen data and can ensure this in practice. The student is able to apply the Pytorch library (or another suitable one) and they can choose an appropriate method for different problems and train it effectively.

Description of prerequisites

Datatieteen aineopintokokonaisuus ja Koneoppiminen–opintojakso (tai vastaavat tiedot). 

Study materials

Dive into Deep Learning: https://d2l.ai/

Completion methods

Method 1

Description:
Ilmoitetaan toteutuskohtaisesti
Evaluation criteria:
Ilmoitetaan toteutuskohtaisesti
Select all marked parts
Parts of the completion methods
x

Participation in teaching (5 cr)

Type:
Participation in teaching
Grading scale:
0-5
Evaluation criteria:
<p>Ilmoitetaan toteutuskohtaisesti</p>
Language:
English, Finnish
Study materials:

Ilmoitetaan toteutuskohtaisesti

Teaching