BEGIN:VCALENDAR VERSION:2.0 PRODID:researchseminars.org CALSCALE:GREGORIAN X-WR-CALNAME:researchseminars.org BEGIN:VEVENT SUMMARY:Paul Atzberger (UC Davis) DTSTART:20200519T230000Z DTEND:20200520T001500Z DTSTAMP:20260423T024445Z UID:MADDD/4 DESCRIPTION:Title: Ge ometric approaches for machine learning in the sciences and engineering\nby Paul Atzberger (UC Davis) as part of Mathematics of Data and Decisio ns @ Davis\n\n\nAbstract\nThere has been a lot of interest recently in lev eraging machine learning approaches for modeling and analysis in the scien ces and engineering. This poses significant challenges and requirements r elated to data efficiency\, interpretability\, and robustness. For scient ific problems there is often a lot of prior knowledge about general underl ying physical principles\, existence of low dimensional latent structures\ , or groups of invariances or equivariances. We discuss approaches for re presenting some of this knowledge to enhance learning methods by using res ults on manifold embeddings\, stochastic processes within manifolds\, and harmonic analysis. We show how the approaches can be used for high-dimens ional stochastic dynamical systems with slow-fast time-scale separations t o learn from observations\, slow variable representations and reduced mode ls for the dynamics. We also discuss a few other examples where utilizing geometric structure has the potential to improve outcomes in scientific m achine learning.\n LOCATION:https://researchseminars.org/talk/MADDD/4/ END:VEVENT END:VCALENDAR