Duration
30h Th, 5h Pr, 45h Proj.
Number of credits
Lecturer
Language(s) of instruction
English language
Organisation and examination
Teaching in the second semester
Schedule
Units courses prerequisite and corequisite
Prerequisite or corequisite units are presented within each program
Learning unit contents
Latest developments in deep machine learning have enabled great and unprecedented advances in systems for visual recognition, speech and text understanding or autonomous intelligent agents. In this context, this course is a deep dive into the details of deep learning architectures, with a focus on learning end-to-end models for these tasks. Students will learn to implement, train and debug their own neural networks and gain a detailed understanding of cutting-edge research in the field. The course will also tour recent innovations in inference methods, including differential inference, adversarial training and Bayesian deep learning.
Topics to be covered (tentative and subject to change):
- Fundamentals of machine learning
- Neural networks
- Convolutional neural networks
- Training neural networks
- Recurrent neural networks
- Auto-encoders and generative models
- Generative adversarial networks
- Uncertainty
- Adversarial attacks and defenses
Learning outcomes of the learning unit
At the end of the course, the student will have acquired a solid and detailed understanding of the field of deep learning. He/she will have studied both well-established and novel algorithms (in theory and in practice), and will also have become familiar with some of the many open research questions and challenges of the field.
Prerequisite knowledge and skills
Programming experience. Probability and statistics.
Following "ELEN0062 Introduction to Machine learning" before taking this class is strongly recommend.
Planned learning activities and teaching methods
- Theoretical lectures
- Programming tutorials
- Reading assignment
- Programming project
Mode of delivery (face-to-face ; distance-learning)
Lectures will taught face-to-face. Projects will be carried out remotely.
Recommended or required readings
Slides will be made publicly available on GitHub during the semester.
Assessment methods and criteria
- Exam (50%)
- Reading assignment (10%)
- Programming project (40%)
Work placement(s)
Organizational remarks
The website for the course is https://github.com/glouppe/info8010-deep-learning
Contacts
- Teacher: Prof. Gilles Louppe (g.louppe@uliege.be)
- Assistants: Antoine Wehenkel, Matthia Sabatelli.
Adaptation of teaching commitments following the COVID-19 pandemic for the May-June 2020 session
Teaching methods implemented : distance-learning
Lectures and tutorials are broadcast live on Twitch and then available as podcasts.
Assessment subjects
The oral exam that should have evaluated the study and understanding of the theoretical content is cancelled.
Instead, the evaluation will be based on the reading assignment and on the project only. Therefore, the focus of the evaluation will be to assess whether the student is able to implement and use Deep Learning models for solving a practical problem. The knowledge and understanding of the theoretical foundations of the course will be minimally evaluated through the project report and the oral defense of the project.
Assessment methods
- The oral exam is cancelled. - Reading assignment (accounting for 20% of the grade, instead of 10%) - Programming project, including a written report and a remote oral defense on Lifesize (accounting for 80% of the grade, instead of 40%)
Contacts
Adaptation of teaching commitments following the COVID-19 pandemic for the Aug-Sept 2020 session
Assessment subjects
Same as for June 2020.
Assessment methods
Same as for June 2020.