Description of the Topics of the Area Cluster
The topic of this cluster is machine learning, including deep learning.
It covers, e.g., representation learning and grounding; sequential decision-making and reinforcement learning; learning from small data sets, GANs and incremental learning; multi-task and transfer learning; and explainable learning.
The cluster collects research topics covering the following WASP keywords:
data-driven decision making, deep learning, explainable machine learning, generative models, machine learning theory, natural language processing, neuro-sybmolic AI, reinforcement learning, representation learning, probabilistic models, statistical learning theory, topological data analysis, transfer, low-shot, semi- and un-supervised learning
Past Cluster Activities
- Cluster kick-off (during winter conference 2022)
Future Cluster Activities
- Cluster seminar and cluster meeting
What: Using the Berzelius SuperPOD by Johan Raber.
Date: April 28, 2022 at 13:30.
Location: Zoom
Abstract:
An introduction to good practises using the Berzelius SuperPOD is presented through slides and live demonstrations on the system. The overarching theme for the presentation is the efficient use of compute time granted on the system. These topics will be covered in particular:- Tools for job monitoring and control
- Moving to multi-node jobs
- Data set placement and formatting
- Setting up and managing your user environment
- Working reproducibly: linux containers
- Study visit in AmsterdamWhat: Study visit about geometric machine learning.
Date: May 31 – June 3
Location: Amsterdam
Program: site visits at Qualcomm, UvA, more to come - Regular Zoom meetings (times to be announced separately)
- Site visits (to be announced separately)
Cluster Resources
- Slack channel: #area_ml in wasp-sweden. You need to be part of the slack organization wasp-sweden to access the channel. If you joined the cluster via the regular process, you should get invited to our channel automatically. We use the slack channel to share information and collect feedback. We encourage active participation.
- Mailing list: Area_ML@wasp-sweden.se. We use the mailing list to disseminate important information about the cluster and cluster activities. All cluster members can send mails to the mailing list.
Contact all Members
E-mail address to all cluster members: Area_ML@wasp-sweden.se