The course is worth 3 points of credit, a project credit and an English course credit, all lectures and assignments would be given in English only. Assignment submissions are in pairs only.
Deep Learning is a transformative technology at the forefront of AI, providing essential tools to innovate across diverse domains—from medical diagnosis to autonomous driving. This graduate-level course provides a comprehensive foundation in the theory and algorithms behind deep learning, along with efficient software techniques for training models.The course will cover the following key topics: multilayer perceptrons, convolutional networks, recurrent networks, unsupervised training, optimization methods, transformers, and generative models.
Upon completing this course, the student will be able to:
- Construct deep neural networks, understanding both the theoretical and practical design aspects.
- Apply optimization methods to effectively train deep neural networks and solve other related optimization tasks.
- Implement and run deep learning applications on hardware accelerators, using both scratch implementations and ready-made frameworks.
- Start research projects related to the field.