News | Jan Povala

June 12th, 2023: I presented our work at the Turing on leveraging wind measurements from airborne airfraft for tactical air traffic control at the UNCECOMP 2023 conference in Athens, Greece. Slides available here.
March 21, 2023: Our paper with Lara Vomfell on understanding spillovers in domestic abuse reporting has been accepted at the Journal of the Royal Statistical Society. Series A (Statistics in Society). See the paper here.
June 7th, 2022: I presented our CMAME Paper on Variational Bayes for Inverse Problems at the ECCOMAS 2022 Congress. Slides available here.
April 22, 2022: I presented our CMAME Paper on Variational Bayes for Inverse Problems at Φ-ML meets Engineering series. Slides available here.
Ferburary 1, 2022: Our paper on using Variational Bayes for Inverse Problems involving elliptic PDEs has been accepted in Computer Methods in Applied Mechanics and Engineering. A big thanks to the co-authors for guidance.
September 21, 2021: I presented our work on Variational Inference for PDEs fin a poster at the Cambridge Centre for Smart Infrastructure and Construction.
August 19, 2020: I presented a poster for our paper on spatial Poisson mixtures at the SMILES Summer School in Machine Learning 2020 (poster) (slides).
June 12, 2020: Our paper on spatial Poisson mixtures applied to London burglary crime has been accepted at the Journal of the Royal Statistical Society. Series C (Applied Statistics).
February 25, 2020: I presented our work on spatial mixture models at the Hierarchical Modelling Workshop held at Cambridge Centre for Data-Driven Discovery. Slides available here.
February 19, 2020: I presented our work on spatial mixture models applied to London burglary crime at the Cambridge-Imperial Computational PhD Seminar. Slides available here.
January 2019: I spent six weeks with the Strategic Insights Unit at the Metropolitan Police Service in London ( Imperial article).
July 2018: I presented my MRes thesis work to the ICONIC advisory board. Slides available here (Adobe Reader is required for animations to work).
May 2018: ICONIC meeting presentation on the log-Gaussian Cox process models inferred using Laplace Approximation. Slides available here (Adobe Reader is required for animations to work).