Machine Learning Working Group
The Machine Learning (ML) Working Group uses machine learning to revolutionize our understanding of the universe.
It meets biweekly on Wednesdays.
Mission Statement
Scope
The Machine Learning Working Group at OKC aims to leverage the transformative potential of machine learning to revolutionize our understanding of the universe and the physical world. Our focus is on developing and applying cutting-edge tools and techniques to effectively manage the vast amounts of data generated by next-generation astronomical, cosmological, and physics experiments. We strive to integrate novel statistical and machine-learning methods to process and interpret complex datasets with increasingly sophisticated models.
How
Our interdisciplinary group fosters collaboration among machine learning experts at OKC, emphasizing areas such as fast deep learning emulators, statistical inference, data clustering, pattern search, and explainable Artificial Intelligence. By staying abreast of the latest advancements in machine learning, we aim to apply these techniques to various physics applications. Through journal clubs, collaborative sessions, invited speaker presentations, and training sessions on GPU utilization and programming languages like JAX, we empower and expand the proficiency of the OKC community in utilizing machine learning techniques.
Specifically, we focus on developing data science techniques relevant to the analysis of upcoming cosmological datasets. By harnessing the power of machine learning, we aim to detect subtle patterns in extensive datasets, classify celestial objects, and explore the behavior of galaxies, dark matter, and cosmic structures. Our ultimate goal is to advance our understanding of the universe through the application of machine learning in cosmology and physics research.
Meetings
Slack channel: #okc-ml
Upcoming meetings of the Machine Learning Working Group
Contact
Justing Alsing
justin.alsing@fysik.su.se
Ludvig Doeser
ludvig.doeser@fysik.su.se
Sinan Deger
sinan.deger@fysik.su.se
Last updated: February 7, 2024
Source: Oskar Klein Centre