- Master course at Radboud University - 6 EC
- Teaching Assistants: Mauricio Diaz-Ortiz, David Ernstberger
This course provides advanced topics in machine learning, with a particular focus on how the statistical physics of disordered systems helps in understanding learning and generalization abilities of neural networks. The course is intended for Master's students in physics and mathematics. This course is the follow-up of CDS: Machine Learning.
- Monte Carlo methods and variational approximation for inference and learning
- Message passing algorithms
- Statistical physics of machine learning: replica method, phase transitions in learning and generalization
- Boltzmann Machines and Deep Belief Networks
- Infinite width limits in deep networks and theory of kernels
- Recurrent networks
- Modern Hopfield Networks, Attention, Transformers
Each lecture comes with an interactive jupyter notebook. Clone the repository to use the notebooks.
- Information Theory, Inference and Learning Algorithms by David MacKay.
- Bayesian Reasoning and Machine Learning by David Barber.