We develop scenario generation methods for risk-averse stochastic programmes. The main idea is to generate more scenarios in parts of the support that correspond to high costs. We do so by means of weighted clustering or importance sampling.
- Speaker
- Date
- Friday 19 Jun 2026, 12:00 - 13:00
- Type
- Seminar
- Room
- ET-14
- Building
- E Building
Candidate solutions are used to assess how high the costs are for the scenarios. In weighted clustering these costs are then used to control for the size of clusters or to change the clustering objective. In importance sampling, the chance of sampling high-cost scenarios is increased. The presented methods can handle general spectral and coherent risk measures (including mean-risk), as opposed to existing methods that can only deal with pure tail risk measures.
About the speaker
Boris Nijhoff is a first-year PhD candidate at the Econometric Institute. His research focuses on developing scenario generation methods for stochastic programmes.
Coordinators
See also
- More information
Lunch will be provided (vegetarian option included).
For more information please contact the Secretariat Econometrics at eb-secr@ese.eur.nl
