Workshops Period III.

June 19, 20, 21

III.5: Information-Based Complexity

Organizers :

Workshop description

The core question of information-based complexity (IBC) is: How many pieces of information are required to solve a (numerical) problem up to a prescribed error tolerance? The problems considered are manifold, including function approximation and learning, numerical integration, optimization, or the solution of PDEs and SDEs. The available information might be given by exact or noisy function values or other types of samples. It is of particular interest how the complexity increases with the dimensionality of the problem (cf. curse of dimensionality versus tractability) and with the desired accuracy (cf. rate of convergence). In view of the recent accomplishments of machine learning, another hot topic in IBC is the question in which situations the power of passive sampling (like iid samples) is comparable to the power of active sampling. We welcome anyone with similar interests to join us for fruitful discussions.


Semi-plenary speakers
Invited speakers

Poster contributors

To be announced.