Lecture Series Computational Science

  • Lecture services

Why this lecture series

The aim of the lecture series is to allow professionals and researchers to quickly get up to speed in some of the current developments in computational science. Each evening, a top-level expert will give an overview of his field and the latest developments therein. A simple dinner and drinks allow the participants to meet new people that can be professionally relevant.

Is this for you?

If you are working in industry, government or a similar organization and want to stay tuned to developments in computational science, this is for you. If you are a researchers in a field that has connections to the topics of the lecture series, this is for you. If you have a general interest in computational science, this is for you. All that is needed is a background on at least HBO level and some previous exposure to computational science.

When and where is this?

The lecture series is given on the following evenings (see details about each evening below):

  • October 2nd Uncertainty Quantification by professor Daan Crommelin (CWI)
  • October 16th Scientific Machine Learning by professor Benjamin Sanders (TU/e)
  • October 30th Multiscale Modelling by professor Roeland Merks (LU)

Each evening lasts from 18:00 until 21:30. It starts with a simple meal, followed by a first lecture of around one hour. After that, there is a coffee break followed by the second lecture. Drinks round up the evening.

Location: Social Impact Factory Vredenburg, Utrecht. A 10-minute walk from Central Station.
Online participation will not be possible.

What are the costs?

The participation fee is € 175, – per evening. If you register for all the evenings, a reduction of 25 euro per evening applies. Reimbursement will be done only if the evening is cancelled by the organization.

How can I register?

Program Details Per Evening

October 2 – Uncertainty Quantification

By Prof. Daan Crommelin (CWI)

Topic of the Lecture

Computational model outputs are influenced by various uncertainties. In this lecture, Prof. Crommelin will cover in short:

  • Uncertainty propagation

  • Surrogate modeling

  • Sensitivity analysis

  • Inverse UQ

The output of computational models can by affected by uncertainties from various sources. Such uncertainties create uncertainties in the computational model output, and it is important to analyze and assess these if the model output is to be used further (e.g. for decision-making).

In this lecture, professor Daan Crommelin will discuss some of the methods that have been developed to analyze and quantify these uncertainties. Topics include uncertainty propagation, surrogate modeling, sensitivity analysis and inverse UQ.

About the Speaker

Daan Crommelin is senior researcher in the Scientific Computing research group of CWI, and holds a part-time position as professor by special appointment of Numerical Analysis and Dynamical Systems at the KdV Institute for Mathematics, University of Amsterdam. He is a member of the CWI management team since 2021 and served as CWI group leader from 2013 until 2021.For more information, visit Daan Crommelin.

October 16 –Scientific Machine Learning

By Prof. Benjamin Sanderse (TU/e)

Topic of the Lecture

Explore how scientific machine learning blends physics-based and data-driven approaches to:

  • Learn laws of nature from data

  • Accelerate scientific discovery

  • Focus on fluid dynamics

In this lecture, we will look at how scientific machine learning can help us discover new science by combining physics-driven and data-driven methods, with an emphasis on fluid flows. We’ll explore how computers can learn the laws of nature from data, and how that helps scientists make faster and smarter discoveries. Come see how machine learning is becoming a powerful new tool in the scientific toolbox.

About the Speaker
Benjamin Sanderse is associate professor in the mathematics department at Eindhoven University of Technology and the group leader of the Scientific Computing group at the Research Institute for Mathematics and Computer Science (CWI). His work focuses on combining physics-based techniques (differential equations) with data-driven techniques (neural networks), also known as Scientific Machine Learning. For more information, visit http://www.thinkingslow.nl

October 30 – Multiscale Modelling

By Prof. Roeland Merks (Leiden University)

Topic of the Lecture

Biological systems are multiscale in nature, requiring:

  • Different models per scale

  • Correct model coupling

  • Efficient computation across many parameters

The hierarchical structure of biological systems makes them excellent examples of multi-scale systems: molecules form signal networks and organelles, these are organized into cells and extracellular structures, the cells together form tissues and organs, etc. Each level arises from the level below, but also influences all the levels below: we call this scale interactions.

A different model is needed for each scale, based on a different type of mathematics. Our challenge is to link these sub-models in a correct way, and to make the combined model efficient enough to calculate it for a large number of parameters.

About the Speaker

Roeland Merks is a professor of Mathematical Biology at Leiden University, where he is affiliated with the Mathematical Institute and the Institute of Biology. He is also the director of the Lorentz Center, a workshop center in Leiden for international scientific meetings. The models he develops help to unravel how complex biological tissues and forms arise from mechanical interactions and chemical communication between cells and their environment. For more information, visit Roeland Merks – Universiteit Leiden.