Academic web page Fredrik Lindsten. https://lindsten.netlify.app/ My warmest congratulations to Dr David Widmann for a well defended PhD thesis, Reliable Uncertainty Quantification in Statistical Learning! Thanks to the opponent, Dino Sejdinovic (University of Adelaide) for the very interesting discussion during the defense, and to the grading committee Meelis Kull (University of Tartu), Henrik Boström (Royal Institute... Wed, 14 Jun 2023 00:00:00 -0700 https://lindsten.netlify.app/2023-06-14-widmann-phd/ https://lindsten.netlify.app/2023-06-14-widmann-phd/ We got a paper published in TMLR on the connection between variational inference and generative flow networks. Heiko Zimmermann, Fredrik Lindsten, Jan-Willem van de Meent, Christian A Naesseth, A Variational Perspective on Generative Flow Networks, TMLR 2023. link This line of work started with a research visit by Heiko to... Mon, 17 Apr 2023 00:00:00 -0700 https://lindsten.netlify.app/2023-04-17-tmlr-paper/ https://lindsten.netlify.app/2023-04-17-tmlr-paper/ I’m looking for two postdocs to join us at the Division of Statistics and Machine Learning at Linköping University. Focus on geometric deep learning for developing new machine learning models and methods to aid in the quest for new ultra-thin materials. Working in a tight collaboration with world leading materials... Tue, 07 Mar 2023 00:00:00 -0800 https://lindsten.netlify.app/2023-03-07-postdocs/ https://lindsten.netlify.app/2023-03-07-postdocs/ Time for another (dry) update… In December last year my PhD student Anna Wigren successfully defended her thesis Sequential Monte Carlo methods for conjugate state-space models. Congratulations Anna! Thanks to the opponent Simon Maskell and the grading committee Harri Lähdesmäki, Víctor Elvira and Tatjana Pavlenko for your excellent review of... Thu, 23 Feb 2023 00:00:00 -0800 https://lindsten.netlify.app/2023-02-23-update/ https://lindsten.netlify.app/2023-02-23-update/ Building on our previous work on Deep Gaussian Markov Random Fields, we – Joel Oskarsson, Per Sidén and I – got a paper accepted for the International Conference on Machine Learning (ICML). We generalize the method from lattice graphs to general graphs (turns out that it was not as straightforward... Fri, 03 Jun 2022 00:00:00 -0700 https://lindsten.netlify.app/2022-06-03-icml-paper/ https://lindsten.netlify.app/2022-06-03-icml-paper/ On Friday 6 May my PhD student Amanda Olmin will defend her licentiate thesis (= roughly half way to PhD), entitled On Uncertainty Quantification in Neural Networks: Ensemble Distillation and Weak Supervision. The questioning will be lead by Hossein Azizpour, KTH Royal Institute of Technology. I’m looking forward to an... Tue, 03 May 2022 00:00:00 -0700 https://lindsten.netlify.app/2022-05-03-olmin-licentiate/ https://lindsten.netlify.app/2022-05-03-olmin-licentiate/ Filip Ekström Kelvinius, Rickard Armiento and I got a paper accepted for Physical Review Materials. The paper is called Graph-based machine learning beyond stable materials and relaxed crystal structures. We investigate the usefulness of graph neural networks, specifically the Crystal Graph Convolutional Neural Network (CGCNN), for predicting formation energies for... Tue, 22 Feb 2022 00:00:00 -0800 https://lindsten.netlify.app/2022-02-22-prm-accepted/ https://lindsten.netlify.app/2022-02-22-prm-accepted/ My PhD student Amanda Olmin and I got a paper accepted for AISTATS 2022! The paper is called Robustness and reliability when training with noisy labels. We investigate the effect of imperfect/noisy labels in the training data when learning a classifier, and in particular how this affect the uncertainty quantification... Thu, 03 Feb 2022 00:00:00 -0800 https://lindsten.netlify.app/2022-02-03-aistats-accepted/ https://lindsten.netlify.app/2022-02-03-aistats-accepted/ We are recruiting several Postdocs in Machine Learning! See this page for details. Wed, 12 Jan 2022 00:00:00 -0800 https://lindsten.netlify.app/2022-01-12-postdocs/ https://lindsten.netlify.app/2022-01-12-postdocs/ I’ve been falling behind with these news updates, so here is a summary over the last six months. I got a few papers accepted for publication at/in ICLR (Calibration tests beyond classification with David Widmann and Dave Zachariah), IJCAI (Likelihood-free Out-of-Distribution Detection with Invertible Generative Models with Amir Ahmadian), IEEE... Tue, 15 Jun 2021 00:00:00 -0700 https://lindsten.netlify.app/2021-06-15-new-papers/ https://lindsten.netlify.app/2021-06-15-new-papers/ We have just opened up two postdoc positions in machine learning for materials disovery at AIICS/STIMA. Additional details are available here. Happy holidays! Thu, 24 Dec 2020 00:00:00 -0800 https://lindsten.netlify.app/2020-12-24-postdocs/ https://lindsten.netlify.app/2020-12-24-postdocs/ Wohoo! I recently got a grant accepted from the Swedish Reserach Council for the project Handling Uncertainty in Machine Learning. The project runs for 4 years. Thu, 03 Dec 2020 00:00:00 -0800 https://lindsten.netlify.app/2020-12-03-vr-grant/ https://lindsten.netlify.app/2020-12-03-vr-grant/ On Oct 28 my PhD student David Widmann successfully defended his licentiate’s thesis (= roughly half way to PhD), entitled Calibration of Probabilistic Predictive Models. The discussion was lead by Meelis Kull, University of Tartu. Wed, 28 Oct 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-10-28-widmann-licentiate/ https://lindsten.netlify.app/2020-10-28-widmann-licentiate/ Congratulations to Johanna Rosén who has been granted 29 MSEK from the KAW foundation for The 2D-Materials Frontier. Together with co-applicant Fredrik Heintz I’m responsible for the machine learning part of the project, where we will use machine learning to aid the search for new and exciting 2D materials. Wed, 07 Oct 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-10-07-materials-project/ https://lindsten.netlify.app/2020-10-07-materials-project/ Paper accepted for NeurIPS, together with Christian A Naesseth and David M Blei! We develop a method for optimizing the “inclusive” Kullback-Leibler divergence KL(p || q) in the context of variational inference by blending VI and MCMC. Check out the paper here. Fri, 02 Oct 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-10-02-neurips-paper/ https://lindsten.netlify.app/2020-10-02-neurips-paper/ My warmest congratulations to Christian A. Naesseth for winning the ISBA Savage Award (Theory and Methods). Mon, 03 Aug 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-08-03-savage/ https://lindsten.netlify.app/2020-08-03-savage/ Paper accepted for MLSP, together with Jakob Lindqvist, Amanda Olmin and Lennart Svensson! We develop a general framework for ensemble distillation that, instead of distilling the ensemble into a single predictive model, retain the distribution over the ensemble members. This provides a more complete description of the predictive uncertainty, making... Tue, 30 Jun 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-06-30-mlsp-paper/ https://lindsten.netlify.app/2020-06-30-mlsp-paper/ Per Sidén and I got a paper accepted for the International Conference on Machine Learning (ICML) P. Sidén and F. Lindsten, Deep Gaussian Markov random fields, ICML 2020. Abstract: Gaussian Markov random fields (GMRFs) are probabilistic graphical models widely used in spatial statistics and related fields to model dependencies over... Tue, 02 Jun 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-06-02-icml-paper/ https://lindsten.netlify.app/2020-06-02-icml-paper/ I’m very happy to accounce my application together with Veoneer AB for a WASP Industrial PhD project got accepted! In August, Hariprasath Govindarajan will join Veoneer and STIMA as an industrial PhD student, working on the project Attentive and Disentangled Representation Learning. Welcome Hari! The company supervisor is Jacob Roll.... Thu, 28 May 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-05-28-wasp-industrial-phd/ https://lindsten.netlify.app/2020-05-28-wasp-industrial-phd/ Today my PhD student Anna Wigren will defend her licentiate’s thesis (= roughly half way to PhD), Exploiting conjugacy in state-space models with sequential Monte Carlo. The external reviewer is Umberto Picchini, Chalmers University of Technology. The presentation is live streamed through Zoom. See this page for additional details. Good... Thu, 14 May 2020 00:00:00 -0700 https://lindsten.netlify.app/2020-05-14-wigren-licentiate/ https://lindsten.netlify.app/2020-05-14-wigren-licentiate/