Login for PhD students/staff at UCPH      Login for others
School on Machine Learning in Physics and Related Sciences
Provider: Faculty of Science

Activity no.: 5934-25-11-13
Enrollment deadline: 01/05/2025
PlaceNiels Bohr Institute, Aud A
Date and time02.06.2025, at: 09:00 - 06.06.2025, at: 16:30
Regular seats60
LecturersTroels Christian Petersen
ECTS credits2.50
Contact personTroels Christian Petersen    E-mail address: petersen@nbi.ku.dk
Enrolment Handling/Course OrganiserTroels Christian Petersen    E-mail address: petersen@nbi.ku.dk
Exam requirementsAttendance along with a final ML-in-science challenge to be handed in (Kaggle style).
Course workload
Course workload categoryHours
Preparation29.00
Lectures10.00
Class Instruction2.00
Practical exercises25.00
Theoretical exercises2.75
Sum68.75
Aim and content
The course will give the student an introduction to knowledge of Machine Learning (ML) and its application in various parts of data analysis in science, physics in particular. The focus will be on application through examples and use of computers.

Formal requirements
A master degree in a science subject, with at least a year’s experience of programming in Python.

Learning outcome
Learning outcome
Knowledge:
• The fundamental methodologies in applying ML to science data.
• The challenges of / solutions to data-MC differences.
• Specific tips & tricks to setting up ML analysis chain.
• Examples of ML usage on various types of data.

Skills:
• Be able to apply ML algorithms to science dataset
• Be able to optimise the ML performance
• Be capable of quantifying and comparing ML performances

Competences:
• Understanding of ML methods along with their pitfalls
• Knowledge of data analysis with ML
• Ability to analyse data using ML in science
• Capability of handling data non-uniformities, along with unbalanced and categorical data.

Target group
Ph.D. students in physics (particle, astro, bio, quantum, etc.) and other Ph.D. students in/with closely related subjects/challenges.

Teaching and learning methods
Each day will have two lectures ( 9-10 and 13-14) followed by practical exercises (10-12 and 14-17).
Two evenings, we will have a mixture of curriculum/social events, which entail class instructions followed by theoretical exercises (i.e. no computers) and then common dinner.
The excursion will (potentially) by a veteran train to Kronborg, which we then visit.

Lecturers
Thea Aarrestad (Fast ML, ++), ETH Zurich
Tilman Plehn (Bayesian NNs, ++), Heidelberg
Troels Petersen (ML experiences/pitfalls), NBI Copenhagen
More (TBA)



Remarks

The School is held in collaboration between the 4EU+ alliance of universities (Prague, Heidelberg, Paris (Panthéon-Assas and Sorbonne), Copenhagen, Geneva, Milan, Warsaw) and the AI-Physics Marie Curie Ph.D. network, but is open to all.
The course will be held in the original Bohr Auditorium at the Niels Bohr Institute on Blegdamsvej in central Copenhagen.

Signing up for the course
Please sign up via website: https://indico.nbi.ku.dk/event/2153/overview

Course fee
PhD students enrolled at universities being af part of the 4EU+ alliance, AIPHY, and for PhD students enrolled at universities that are part of the Danish open market for PhD courses: The course is free of charge

PhD students enrolled at other universities (than listed above): The course fee of 3.000 DKK applies covering tuition, catering during the course and social events.
After you have been accepted for the course we will send you a link to payment of the course fee. 
NB. Payment via creditcard or Mobilepay only.

Participants not enrolled in a university (e.g. post-docs, guest researchers): The course fee of 8.400 DKK applies covering tuition, catering during the course and social events.
After you have been accepted for the course we will send you a link to payment of the course fee. 
NB. Payment via creditcard or Mobilepay only. 

***

All participants: All participants must arrange for travel and accommodation themselves.

 

Search
Click the search button to search Courses.


Course calendar
See which courses you can attend and when
JanFebMarApr
MayJunJulAug
SepOctNovDec



Publication of new courses
All planned PhD courses at the PhD School are visible in the course catalogue. Courses are published regularly.