|
|
| Colloquium for statistics |
|
Speaker:
|
Natalie Neumeyer
|
|
Title:
|
Statistical inference for the error distribution in functional linear models - Show abstract
Abstract: Some recent results on functional linear models with scalar response and functional covariate are presented. In those models it is more challenging to deal with residual-based procedures than in regression models with vector-valued covariates. Typically estimating the functional parameter will dominate the asymptotic distribution of test statistics based on residuals with a slower rate of convergence. As one example we consider a test for independence of errors and covariates based on joint empirical characteristic functionals of residuals and functional covariates. Yet under simple assumptions there are change-point and goodness-of-fit tests for the error distribution with the same asymptotic distribution as for classical linear models. The suggested tests are based on residual empirical distribution functions and residual empirical characteristic functions.
|
|
Schedule:
|
Thu, 19.11.2026 : 14:15 - 15:15 in INF 205, SR 5
|
|
| Colloquium for statistics |
|
Speaker:
|
Victor-Emmanuel Brunel
|
|
Title:
|
Estimation of barycenters in metric spaces with curvature bounded above - Show abstract
Abstract: In non-Euclidean metric spaces, barycenters provide a canonical extension of means: they minimize the expected squared distance to a random point. We study the estimation of the barycenter of a probability distribution from iid data. We work under the assumption that the ambient metric space has curvature bounded above in Alexandrov’s sense: roughly speaking, this means that triangles in that space are no thicker than triangles with the same side lengths on a model space with constant curvature (e.g., hyperbolic plane, Euclidean plane or Euclidean sphere). We consider two estimators: the empirical barycenter of the data and a computationally efficient alternative particularly suited to streaming data. If the data distribution is supported on a small enough domain (e.g., a domain smaller than a hemisphere on a Euclidean sphere), we show that our estimators achieve a parametric rate, with error bounds both in expectation and with high probability. This is based on joint works with Jordan Serres (Université Paris-Sorbonne), Yassine Boukhateb (ENSAE) and Austin Stromme (ENSAE).
|
|
Schedule:
|
Thu, 12.11.2026 : 14:15 - 15:15 in INF 205, SR 5
|
|
| Colloquium for statistics |
|
Speaker:
|
Richard Samworth
|
|
Title:
|
Deep learning with missing data
|
|
Schedule:
|
Thu, 15.10.2026 : 14:15 - 15:15 in INF 205, 5/104
|
|
|