Fast Computing
mathLab @10
Part of the team
Performances optimization
A brief overview
Part of the team
Fracture study
Projects list
Cardiovascular simulations
videos
Part of the team
Cells locomotion study
Shape optimization
Flow optimization
Publications list
Microswimmers locomotion study
Turbulence reconstruction
Flow Studies
Helical swimming in euglenids
Study on plant circumnutations
1 of 20

Community

Since 2010, a laboratory for mathematical modeling and scientific computing devoted to the interactions between mathematics and its applications.

Our blog 

People

A young team of international scientists pursuing frontier research, while expanding the opportunities for a dialogue across academic and disciplinary boundaries.

Join us

 

Master thesis, post-grad/pre-doc and PhD programs

Master thesis projects for graduate students are available under the supervision or co-supervision of SISSA mathLab professors and/or researchers.

SISSA mathLab Research is funded by Mathematics Area of SISSA, European Research Council, European Cooperation in Science and Technology, Italian Ministry for Education, University and Research,  Regional Administration of Friuli Venezia Giulia, European Social Fund, as well as by private and public industries.

Latest events

Neural-Network Interpretability for Time Series Classification Task

Date: 

18/07/2023 - 16:00

Neural networks (NN) have been gaining significant traction for time series classification tasks over the past few years. Yet, they are frequently perceived as black-box tools, whose results may be difficult to interpret. To address this issue, several methods have been proposed to obtain maps of relevance scores highlighting the importance of different time steps for a given model. These methods were initially applied to images, and more recently to time-series data. Yet, interpretability of NN remains challenging. Indeed, interpretability methods typically provide different results, sometimes even diametrically opposite, and may not explain how neurons collaborate to represent specific patterns. In this work, we propose a new evaluation framework for post-hoc interpretability methods applied to time series classification tasks. We argue that this work is a critical step toward understanding NN-based decisions and provide a more robust interpretability workflow. We also present a preliminary study that aims to understand the robustness of the evaluation metrics.

Category: 

Model Order Reduction for Complex Domain Problems in the Time-Continuous Space-Time Setting

Date: 

04/07/2023 - 16:00

Model Order Reduction for Complex Domain Problems in the Time-Continuous Space-Time Setting 

Dr Fabian Key (TU Wien)

Category: 

Subscribe to Front page feed