# Machine Learning 1 (Winter Term 2022/2023)

## Overview

- Course (2/2/0) consisting of:
**Lectures** in TRE/PHYS (Zellescher Weg 16) on Fridays, 09:20–11:10
**Exercise groups**
- online, synchronously, on Wednesdays, 09:20–11:10
- online, asynchronously, through a forum
- in VMB/0302/U on Fridays, 14:50–16:40
- in VMB/0302/U on Fridays, 16:40–18:30

**Self-study and moderated discussion in a forum**
**Final Examination**

- Lecturer: Bjoern Andres
- Teaching Assistant: Shengxian Zhao
**Enrolment (OPAL)**. Additional rules for enrolment may apply, depending on the study programme.
- Creditable toward the modules CMS-CLS-ELG, CMS-CLS-ELV, CMS-COR-MLD, INF-B-510, INF-B-520, INF-BAS2, INF-BAS7, INF-PM-FOR, INF-VERT2, INF-VERT7, MCL-AI, MCL-PI
- A post-exam review will take place in APB-2023 on May 8th, 2023 at 9:00. Registration by email is required two days prior to the event.

## Contents

- Lecture notes
- Exercises
- Lectures
**Introduction** (slides)
**Supervised learning**
- Introduction (slides)
- Learning of disjunctive normal forms (slides)
- Learning of binary decision trees (slides)
- NP-hardness
- Local search algorithm

- Learning of linear functions (slides)
- Logistic regression
- Gradient descent algorithm

**Semi-supervised and unsupervised learning**
- Introduction (slides)
- Partitioning (slides)
- Set partition problem
- Local search algorithms

- Clustering (slides)
- Multicut problem
- Local search algorithms

- Ordering (slides)
- Linear ordering problem
- Local search algorithms

- Classifying (slides)

**Supervised structured learning** (slides)
- Factor graphs
- Conditional graphical models
- Message passing algorithms

## Textbooks