The Swedish e-Science Academy is organised by eSSENCE and is open to anyone interested in e-Science. It is a lunch-to-lunch meeting and this year it will be in Uppsala.

The Swedish e-Science Academy represents a great opportunity to meet fellow e-scientists to discuss common issues and get inspired.

Meals and accommodation for the night between 11-12 October will be provided.

A novelty is that there will also be satellite meetings on the following topics

• Numerical Linear Algebra: In Honour of Maya Neytcheva, 10-11 October, lunch-to-lunch
• InfraVis: National Research Infrastructure for Data Visualization, 11 October, 9-12
• Advancements in Federated Machine Learning, 11 October, 9-12
• Support for HPC Users in Sweden, 12 October,  13-16

Thus, please also consider joining one or two of the satellite meetings.

Preliminary Programme

Wednesday 11 October

Venue: University Building

• 12:00 Lunch (packed lunch for late arrival)
• 13:15 Welcome: Andreas Hellander, eSSENCE Programme Director. David Black-Schaffer, Dean of Research, Faculty of Science and Technology and Chair of the eSSENCE board
• 13:30 Keynote 1: Ola Spjuth, Uppsala University – “How AI and Automation is accelerating drug discovery”
• 14:10 Project presentation 1: Matteo Magnani, Uppsala University
• 14:30 Project presentation 2: Josefin Starkhammar, Lund University
• 14:50 Project presentation 3: Monowar Bhuyan, Umeå University
• 15:10 Poster blitz (1 slide, 45 seconds per poster)
• 15:40 Poster session with refreshments & snacks
• 17:30 Time to check in at the Hotel
• 19:00 Dinner at Hambergs festvåningar

Thursday 12 October

Venue: Room 80101, Ångström Laboratory

• 09:00 Keynote 2
• 09:45 Coffee/tea
• 10.00 Project presentation 4: TBA
• 10:20 Project presentation 5: Jesper Wallentin, Lund University
• 10:40 Project presentation 6: Martin Berggren, Umeå University
• 11:00 Panel discussion: Future needs for e-infrastructure for e-Science
• 12.00 Closing words
• 12:05 Farewell and Thank You!
• Lunch

Accommodation

There are 50 rooms booked at Hotel Scandic Uplandia in central Uppsala for the night between 11 and 12 October. To claim a room, just select "I want accommodation ..." in the registration form, but since rooms are limited, register for the Swedish e-Science Academy as soon as possible to avoid missing out.

Registration

To register for the Swedish e-Science Academy, including meals and one night accommodation 11-12 October, please use the registration form (Google).

Satellite Meetings

Numerical Linear Algebra: In Honour of Maya Neytcheva

Time: 10-11 October

Venue: Häggsalen (Room 10132), Ångström Laboratory

In honor of Professor Maya Neytcheva, Uppsala University, who is retiring in 2023 after a long and successful career, we hold this satellite workshop in the subject of numerical linear algebra. Hence, the purpose is the celebration of the scientific work of Professor Maya Neytcheva, the sharing of current research, and fostering future research in the field.

The workshop consist of presentations and discussions related in a wide sense to the subject area of Maya Neytcheva’s research.

Programme

Tuesday 10 October

• 13:00 – 13:15 Introduction
• 13:15 – 14:00 Talk 1.1
• 14:15 – 15:00 Talk 1.2
• 15:15 – 16:00 Talk 1.3
• 16:15 – 17:00 Talk 1.4
• 18:30 Dinner

Wednesday 11 October

• 09:15 – 10:00 Talk 2.1
• 10:15 – 11:00 Talk 2.2
• 11:15 – 12:00 Talk 2.3

Confirmed speakers

• Michal Béreš, Technical University of Ostrava
• Paolo Bientinesi, Umeå University
• Petia Boyanova, Bulgarian Academy of Sciences
• Philipp Birken, Lund University
• Ivo Dravins Ruhr, Universität Bochum
• Roman Iakymchuk, Uppsala University
• Elias Jarlebring, Royal Institute of Technology
• Svetozar Margenov, Bulgarian Academy of Sciences
• Stefano Serra-Capizzano, University of Insubria
• Eddie Wadbro, Karlstad University

Registration

Registration link for Numerical Linear Algebra: In Honour of Maya Neytcheva (and dinner)

InfraVis: National Research Infrastructure for Data Visualization

Time: 11 October, 9-12

Venue: Theatrum Visuale (room 100155, entrance floor), Ångström Laboratory

InfraVis is a national research infrastructure providing support for visualisation of data from all disciplines.

Programme

• 09:00 – 09:30 General presentation of InfraVis (Ingela Nyström, UU)
• 09:30 – 10:00 Visualization in the Swedish HPC landscape (Mario Romero, KTH)
• 10:00 – 10:15 Mingle and fika 
• 10:15 – 11:00 Dimensionality Reduction Techniques and Visualization (Anders Hast, UU)
• 11:00 – 11:30 A couple of scientific cases having benefitted from InfraVis support (InfraVis Application Experts)
• 11:30 – 12:00 Visual Text Analytics (Andreas Kerren, LnU/LiU) 

Registration

Registration link for Infravis satellite meeting

The registration deadline is Friday 29 September.

Advancements in Federated Machine Learning

Time: 11 October, 9-12

Venue: Polhemssalen (Room 10134), Ångström Laboratory + Zoom

Federated machine learning (FedML) has emerged as a unique approach for privacy-preserving data analysis, enabling machine learning in scenarios where centralising data is impractical. By allowing participants to work locally on their private datasets and contribute to a joint federated global model, FedML offers the potential to train models while simultaneously complying with legislation such as the GDPR. It empowers advanced machine learning even with private, sensitive, or hard-to-move data, thereby marking a paradigm shift in the field.

The objective of this workshop is to establish a collaborative platform for Swedish researchers to share their expertise and foster cooperation on mutually interesting ideas related to FedML. By facilitating engaging discussions and promoting cooperation among researchers, the workshop aims to drive advancements in federated machine learning.

Programme

• 09:00 – 09:15 Welcome
• 09:15 – 09:35 On the Role of Federated Learning with 6G on the Horizon (Andreas Johnsson, Ericsson Research)
• 09.35 – 10.00 Regulating Data Privacy – Balancing Legal and Technical Challenges (Sandra Friberg, Dept. of Law, Uppsala)
• 10.00 – 10:30 Coffee break 
• 10.30 – 10.55 Findings from the MELLODDY project: federated learning for molecule property prediction (Lewis Mervin, Astra Zeneca)
• 10.55 – 11.20 Federated Learning: A Panacea in Medical Applications of Machine Learning? (Tommy Löfstedt, Dept. of Computing Science, Umeå)
• 11:20 –12.00 Open Discussion Session (Convener: Salman Toor, Dept. of Information Technology, Uppsala and Johan Östman, AI Sweden)

Registration

Registration link for Advancements in Federated Machine Learning

Additional information

Additional information, including the Zoom link, is available as a PDF document (52 kb).

Support for HPC Users in Sweden

Time: 12 October,  13-16

With the resent transition from SNIC to NAISS there is an uncertainty as to how the need for user support for HPC users in Sweden will be met. It is anticipated that the new organisation with branches will be formed early in the fall of 2023. Thus, it is of interest to discuss at a high level what needs are to be met, how the organisation for exchange of competences can be arranged, and what the needs and capabilities of the HEI:s are. In this workshop the status of Swedish user support for HPC will be discussed, and future scenarios will be mapped. The aim of the workshop is to establish a common understanding and a manifest outlining possible solutions to the challenges in the current and future HPC landscape in Sweden.

Programme

• 13.00 – 13.45 General introduction of the current state of HPC user support and the road ahead. Introduction group discussions.
• 13.45 – 14.15 Group discussions on the needs and capabilities
• 14.15 – 14.30 Presentation of the group discussions. Introduction to future challenges and opportunities.
• 14.30 – 15.00 Group discussions on future challenges and opportunities
• 15.00 – 15.15 Presentation of the group discussions
• 15.15 – 16.00 Summing up and authoring of draft “manifest”

Registration

Registration link for Support for HPC Users in Sweden