S
Sebastian M. Schmon
Researcher at University of Oxford
Publications - 19
Citations - 171
Sebastian M. Schmon is an academic researcher from University of Oxford. The author has contributed to research in topics: Computer science & Bayesian probability. The author has an hindex of 4, co-authored 10 publications receiving 61 citations.
Papers
More filters
Journal ArticleDOI
AnoDDPM: Anomaly Detection with Denoising Diffusion Probabilistic Models using Simplex Noise
TL;DR: A novel partial diffusion anomaly detection strategy that scales to high-resolution imagery, named AnoDDPM, and a multi-scale simplex noise diffusion process that gives control over the target anomaly size is developed.
Journal ArticleDOI
Large-sample asymptotics of the pseudo-marginal method
TL;DR: A study of this limiting chain allows us to provide parameter dimension-dependent guidelines on how to optimally scale a normal random walk proposal, and the number of Monte Carlo samples for the pseudo-marginal method in the large-sample regime.
Journal ArticleDOI
Estimating the density of ethnic minorities and aged people in Berlin: multivariate kernel density estimation applied to sensitive georeferenced administrative data protected via measurement error
TL;DR: This work proposes multivariate non-parametric kernel density estimation that reverses the rounding process by using a Bayesian measurement error model, applied to the Berlin register of residents for deriving density estimates of ethnic minorities and aged people.
Journal ArticleDOI
Investigating the Impact of Model Misspecification in Neural Simulation-based Inference
TL;DR: It is concluded that new approaches are required to address model misspecification if neural SBI algorithms are to be relied upon to derive accurate conclusions about stochastic simulation models.
Posted Content
Generalized Posteriors in Approximate Bayesian Computation
TL;DR: First, the accept/reject step in ABC is re-interpreted as an implicitly defined error model, and it is argued that these implicit error models will invariably be misspecified.