Skip to content
Technische Universität Berlin Deutsch
Course catalog and module transfer system

Workshop Machine Learning Foundations for PhysicistsModule description #41169 / Version 1

← Module page Deutsche Fassung
This module description is not valid in the current semester (Winter semester 2026/27).
Module / Version#41169 / #1
ValiditySummer semester 2024 – Winter semester 2024/25
Available language(s)English
Credit points9
Person responsible for the moduleMontavon, Gregoire
GradingUngraded
Exam formatPortfolio examination
Language(s) of instructionEnglish

Affiliation

FacultyFakultät IV
InstituteInstitut für Softwaretechnik und Theoretische Informatik
Academic chair34352000 FG Maschinelles Lernen

Contact

OfficeMAR 4-1
Contact personNakajima, Shinichi
E-mail addressnakajima@tu-berlin.de
Websitehttps://web.ml.tu-berlin.de/

Learning outcomes

This workshop is designed to give students with a background in physics an introduction to the field of machine learning, as well as to introduce them to state-of-the-art research in machine learning methods in physics.
Students will have a good overview of machine learning methods applied in physics, such as accelerated simulations of lattice gauge theories, quantum, classical and quantum-inspired machine learning, and computation of generalized parton distributions with learnable denoising approaches.

Content

- Foundations in Machine Learning:
-> Classical Methods like Linear Regression and the Perceptron
-> Unsupervised Learning: Clustering
-> Bayesian Methods
-> Introduction Artificial Neural Networks, Deep Learning, Autodiff Frameworks like Pytorch.
-> Specialized Network architectures, like Recurrent and Convolutional Neural Networks
-> Geometric Deep Learning
-> Deep Generative Models, like Normalizing Flows

Module components

Compulsory area

The following courses are mandatory for this module:

CoursesTypeNumberFrequencyLanguageHours per weekCourse catalog
Beginners Workshop Machine LearningProject Winter semester–6 ISIS

Workload and credit points

Beginners Workshop Machine Learning (Project)

Workload descriptionMultiplierHoursTotal
Attendance158 h120 h
Pre/post processing156 h90 h
Group Work410 h40 h
250 h

The workload of the module adds up to 250 hours. The module therefore comprises 9 credit points.

Description of teaching and learning methods

Students will have an intensive block workshop, which includes lectures, presentations from experts in related fields and different exercises.
There will be minor exercises and bigger group exercises which will be used to measure the performance of the students.

Requirements for participation and examination

Desirable prerequisites for participation in the courses:

A strong background in physics is strongly recommended.
This course will be offered in English. Thus, strong English language skills (C1) are strongly recommended.

Mandatory requirements for the module examination registration:

This module has no examination requirements.

Completion of the module

GradingUngraded
Exam formatPortfolio examination
Type of portfolio examination100 points in total
LanguageEnglish

Test elements

NamePoints/WeightCategoryDuration/Extent
(Deliverable assessment) Group Exercise 125practical2-5 pages
(Deliverable assessment) Group Exercise 225practical2-5 pages
(Deliverable assessment) Group Exercise 325practical2-5 pages
(Deliverable assessment) Group Exercise 425practical2-5 pages

Exam description

The students will be evaluated based on the four given group exercises. Which consist of implementing ML models to various problems and documenting the work.
The first two exercises will tackle foundations of Machine Learning, the latter two applications of ML algorithms to physical problems introduced during the workshop.

Duration of the module

The following number of semesters is estimated for taking and completing the module: 1 semester(s).

This module can be started in the following semesters: Winter semester.

Maximum number of participants

The maximum capacity of participants is 18.

Registration procedures

Registration for the module is handled through BIFOLD.

Recommended reading, lecture notes

Lecture notes

unavailable

Electronic lecture notes

unavailable

Recommended literature

No recommended literature given

Assigned programs

This module is not used in any program.

Miscellaneous

Not specified