The most reliable way of contacting me: andrey.lokhov (at) gmail.com
Open positions: check the positions anouncement page
News and events:
- Don’t miss the third edition of the Physics Informed Machine Learning (PIML 2020) workshop that I co-organize in Santa Fe, NM on January 13-17, 2020. Apply for travel grants before November 15, 2o19! Previously, we organized and hosted PIML 2018 and PIML 2016.
- I am co-organizing an IMA workshop “Theory and Algorithms in Graph-Based Learning” to be held September 14-18 2020 in Minneapolis, MN.
- Our demonstration of real-time anomaly detection and classification in streaming data from power grids has been presented at NeurIPS 2019.
- Our new paper on learning arbitrary discrete graphical models is out on arXiv.
My research interests lie at the intersection of statistical physics, computer science and machine learning. I develop new statistical physics inspired algorithms to deal with hard problems in graphical models, machine learning, complex systems, quantum computing and data mining. I am also interested in general optimization, inverse, combinatorial and out-of-equilibrium problems on graphs. Check out the research higlights for details.
2015-2017 Postdoctoral fellow in the Center for Nonlinear Studies, Los Alamos National Laboratory, Los Alamos, United States
2011-2014 PhD in statistical physics at Laboratoire de Physique Théorique et Modèles Statistiques, Université Paris-Sud, Orsay, France
PhD thesis: Dynamic cavity method and problems on graphs, under supervision of Marc Mézard, defended on November 14, 2014
2010-2011 Master in theoretical physics at M2 CFP Physique Théorique, École Normale Supérieure, Paris, France
Master thesis: Physics of heavy flavors in QCD, under supervision of Gregory Korchemsky, defended in September 2011
2008-2011 Undergraduate studies at École Polytechnique, Palaiseau, France
2005-2011 Undergraduate studies at Novosibirsk State University, Novosibirsk, Russia