This course offers a graduate-level exploration of the mathematical and algorithmic foundations of machine learning, with an emphasis on deep neural networks. Students will study optimization theory, generalization bounds, and architecture design before advancing to contemporary topics such as transformers, diffusion models, and reinforcement learning from human feedback.
ML Foundations
Bias-variance tradeoff, hypothesis space, evaluation metrics, MLE, MAP estimation, probability distributions, KL divergence & information-theoretic measures
Neural Networks & Optimization
FFNs, expressivity, activation functions, SGD, momentum, Nesterov, Adam, LR scheduling, gradient clipping
Architectures
CNNs, RNNs, LSTMs, Attention, Transformers, SSMs, xLSTMs
Generative Models
VAEs, GANs, Diffusion Models, LLMs
| Week | Date | Topic | Type | Materials |
|---|---|---|---|---|
| 01 | 07.10.2025 | Basics of Machine Learning — Part I Neural networks overview, reinforcement learning, sequential models, generative models |
Lecture | |
| 01 | 09.10.2025 | Basics of Machine Learning — Part II Bias-variance tradeoff, hypothesis space, evaluation metrics |
Lecture | |
| 02 | 14.10.2025 | Statistical Approaches in ML — Part I Estimators, MLE, MAP, distributions |
Lecture | |
| 03 | 28.10.2025 | Statistical Approaches in ML — Part II Divergences |
Lecture | |
| 03 | 28.10.2025 | Assignment I — Statistical Learning Foundations |
Assignment | |
| 03 | 28.10.2025 | Neural Networks — Part I FFNs, expressivity, activation functions, gradient descent |
Lecture | |
| 03 | 30.10.2025 | Neural Networks — Part II Gradient descent, neural network architecture |
Lecture | |
| 04 | 04.11.2025 | Gradient Descent Methods GD, SGD, momentum-based GD, Nesterov-accelerated GD, LR scheduling, adaptive LR methods, gradient clipping |
Lecture | |
| 05 | 06.11.2025 | Convolutional Neural Networks — Part I Convolution operator, kernels, CNNs, hierarchy |
Lecture | |
| 05 | 11.11.2025 | Convolutional Neural Networks — Part II LeNet, ImageNet, AlexNet, ZFNet, VGGNet, ResNet, guided backpropagation |
Lecture | |
| 06 | 18.11.2025 | Sequential Modelling Introduction, RNNs, LSTMs, GRU |
Lecture | |
| 06 | 20.11.2025 | Advances in Sequential Modelling Seq2Seq, xLSTM, SSMs, Mamba |
Lecture | |
| 06 | 20.11.2025 | Assignment II — Neural Networks and Sequential Models |
Assignment | |
| 07 | 27.11.2025 | Transformers — Part I Attention Mechanism, Transformers |
Lecture | |
| 07 | 04.12.2025 | Transformers — Part II Attention decoder, language modelling, GPT family, LoRA, RLHF, vision transformers |
Lecture | |
| 08 | 08.12.2025 | Generative Modelling — Part I Generative models, autoencoders, variational autoencoders, β-VAEs, masked autoencoders |
Lecture | |
| 08 | 08.12.2025 | Assignment III — Variational Autoencoder |
Assignment | |
| 08 | 18.12.2025 | Generative Modelling — Part II GANs, mode collapse, conditional GANs, DCGANs, applications |
Lecture | |
| 09 | 15.01.2026 | Diffusion Models Forward and reverse diffusion, training, applications |
Lecture |
(best: 1)
Source: Official university course feedback system, WS 2025/26.