Summary¶
These are the teaching materials for the course Aprendizaje Profundo (deep learning), coordinated by Antonio Pertusa and also taught by professors Juan Antonio Pérez, Andrés Fuster, Jorge Azorín, and Jorge Calvo.
For information regarding the course assessment, please refer to the Moodle contents at UACloud and the course official info page.
Methodology and evaluation¶
- The theory contents should be revised by the student before attending the practical sessions. A test will be answered to ensure that these contents are read and understood by the student.
- Practical assignments (jupyter notebooks) are to be completed as indicated in each assignment description. Each of the course blocks will have one or more practical assignments.
The theory tests represent 30% of the final grade. The assignments represent 70% of the final grade.
Schedule¶
The course has the following blocks:
- Feb. 6: From Shallow to Deep Neural Networks (Andrés Fuster)
- Feb. 6 & Feb. 13: Convolutional Neural Networks (Jorge Azorín)
- Feb. 13: Deep Reinforcement Learning (1/2) (Jorge Calvo)
- Feb. 20: Recurrent architectures (Juan Antonio Pérez)
- Feb. 20: Fine tuning and knowledge distillation (Juan Antonio Pérez)
- Feb. 27: Few and Zero Shot Learning (Antonio Pertusa)
- Feb. 27: Self-Supervised Learning (Antonio Pertusa)
- Mar. 6: Deep Reinforcement Learning (2/2) (Jorge Calvo)
More information¶
The source code of these pages, written in markdown for MkDocs, is available on GitHub.
You can obtain a local copy of these pages (e.g., for offline access) by executing:
wget --mirror --no-parent --convert-links --page-requisites https://pertusa.github.io/ap
Please note that the content may change throughout the course.