- Part 1 - Introduction
- 1.1: Introduction: machine learning, a couple of applications [46 slides]
- 1.2: A short history of deep learning [10 slides]
- 1.3: An extremely short crash course on fitting and generalization [23 slides]
- Part 2 - Neural Networks and PyTorch
- 2.1: Frameworks and Tensors {+PyTorch} [25 slides]
- 2.2: (Generalized) Linear Models {+PyTorch}
[34 slides]
Code: [linearreg.py]
Code: [logreg.py]
Data: [sand_slope.csv]
Data: [breast-cancer-wisconsin-cleaned.csv] - 2.3: Multi layer models + Universal approximation theorem {+PyTorch}
[32 slides]
Code: [mlp.py]
Data: [MNIST-png.zip] - 2.4: Train/Validation/Test split; Tensorboard {+PyTorch}
[14 slides]
Code: [board.py] - 2.5: Gradient Backpropagation & Autograd {+PyTorch} [17 slides]
- 2.6: Stochastic Gradient Descent (SGD) and Variants (ADAM, RMSProp) {+PyTorch}
[15 slides]
Code: [learningrates.py] - Part 3 - Machine Learning for Control
- 3.1: Imitation Learning {+PyTorch}
[15 slides]
Code: [cartpol_test.py]
Code: [beh_cloning.py] - 3.2: Reinforcement Learning [25 slides]
- Part 4 - Architectures beyond MLPs
- 4.1: Convolutions and shift-equivariance {+PyTorch}
[45 slides]
Code: [lenet.py] - 4.2: Recurrent Neural Networks and Time Series [25 slides]
- 4.3: Transformers {+PyTorch} [15 slides]