ʵ

News

CEST group wins supercomputing Grand Challenge

Researchers from the CEST group have successfully applied for a special supercomputing assignment to carry out research related to climate change.
An image showing an arrangement of data servers from supercomputer Mahti
Supercomputer Mahti, BullSequana XH2000 from Atos. Image: CSC 2020.

CEST group researchers from the Department of Applied Physics were awarded a competitive Grand Challenge project on Finland’s biggest supercomputer Mahti to explore “Active machine learning methods for atmospheric science applications”.

Milica Todorović and Patrick Rinke will manage a 15 million core-hours project, awarded to advance cutting-edge research with large computational needs, in cross-disciplinary collaboration with the groups of Dr. Theo Kurtén and Prof. Hanna Vehkamäki at University of Helsinki.

Active learning dataset curation techniques, developed by Kunal Ghosh at CEST, will be employed to optimize the quality of machine learning of aerosol condensation properties for small atmospheric molecules. This is important information for atmospheric climate models that shape public policy on climate change.

  • Updated:
  • Published:
Share
URL copied!

Read more news

Log2Motion simulation with a musculoskeletal model using a smartphone.
Press releases, Research & Art Published:

Tired of swiping? Now an AI simulation helps us understand why

Screen logging tells us where smart phone users tap and swipe, but now researchers have developed a musculoskeletal model that helps understand the physical effort that goes into these motions
Grönlund-Nisunen's piece Insight as seen from below
Appointments, Research & Art Published:

New professors appointed to the School of Arts, Design and Architecture

A summary of all new professors appointed since September 2024.
Primary pupils sit spaced at wooden desks in a bright classroom, facing the teacher at the front.
Cooperation, Research & Art Published:

The Educational Partnership project is moving forward in Espoo – cooperation between guardians and schools is being developed through participatory methods

The two-year project explores and develops cooperation between guardians and schools using service design methods.
Person sits by a glowing screen with pixel smile and floating heart chat bubbles between them
Press releases, Research & Art Published:

AI companions can comfort lonely users but may deepen distress over time

Long-term use of AI companions may give comfort, but research indicates it may negatively impact users’ wellbeing and their ability to navigate real world relationships.