These lectures were given in Jan. 2022 to a class of PhD students in Physics at Bicocca University in Milan. Although they have a Physics flavour, they can also be followed by other students with different backgrounds.

Here are my contacts. Consider to follow me if you appreciate my contents!

• Twitter (I post about AI/ML/DL and Physics)
• LinkedIn (same as twitter, but with a slightly more serious profile pic)

By the way, a new session of the School of AI (it is free!) at Pi School is starting soon. For two intense months, you will work on a real-world challenges, supervised by experts. Here are more info and how to apply.

Lectures

1. Deep Learning Theory and PyTorch Introduction
2. Build a Custom Multilayer Perceptron from Scratch
3. A Complete Deep Learning Training, Evaluation and Test Pipeline (with Hyper-params tuning and k-fold cross validation)
4. From Dense to Convolutional Layers
5. The Variational Auto-encoder
6. Deep Convolutional Generative Adversarial Networks

References

Books (must have)

• Deep Learning - Goodfellow, Bengio, Courville
• Information Theory, Inference, and Learning Algorithms - David J. C. Mackay
• Deep Learning with Python - Fancois Chollet
• Reinforcement Learning - Sutton, Barto

Popular Books (Information Theory)