Machine Learning 1 – Master's Course

Technische Universität Berlin · Prof. Klaus-Robert Müller

This repository contains coursework from the graduate-level Machine Learning 1 course taught by Prof. Dr. Klaus-Robert Müller at Technische Universität Berlin during the Winter Semester 2021/22. The course is part of the Master’s in Computer Science program and explores how machines can learn from data — from statistical foundations to modern AI techniques.

Assignments include both analytical derivations and programming tasks.

---

Topics Covered

The course introduced a wide range of core machine learning methods, including dimensionality reduction, classification, model selection, neural networks, ensemble learning, clustering, and explainable AI:

1. Bayes Decision Theory – Decision boundaries, loss functions, Gaussian classifiers
2. Parameter Estimation – maximum likelihood estimation, bias–variance trade-off
3. Principal Component Analysis (PCA) – Dimensionality reduction, eigendecomposition
4. Fisher Discriminant – Linear classification, between/within-class scatter
5. Model Selection – Overfitting, cross-validation, regularization
6. Neural Networks 1 – Perceptrons, backpropagation, gradient descent
7. Learning Theory & Kernels – Vapnik–Chervonenkis dimension, margins, Mercer's theorem
8. Support Vector Machines (SVMs) – Dual optimization, hard/soft margin
9. Kernel Ridge Regression – Kernel trick, regularized least squares
10. Boosting – AdaBoost, ensemble methods
11. Decision Trees and Random Forests – Entropy, pruning, bagging
12. Neural Networks 2 – Deep learning architectures
13. Latent Variable Models / Clustering – K-means, expectation–maximization, mixture models
14. Explainable AI (XAI) – Feature attribution, interpretability

---

Contents

The repository is organized into weekly folders (e.g., Week01, Week02, ...) following the course progression. Each folder typically contains:

• Analytical_Homework.pdf files with analytic derivations, proofs, and handwritten solutions
• Programming_Homework.ipynb files with programming assignments (Jupyter Notebooks)

This coursework is based on my own work, collaborative discussions within my homework group, and public class materials. Redistribution or reuse of these materials for educational or institutional use is not permitted.

---

👉 Also see: Machine Learning 2 – Master's Course