BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//jEvents 2.0 for Joomla//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:606vabbivnjtf8s1pq4tnhm8c8@google.com
CATEGORIES:RTportal.ru
CREATED:20170606T092329
SUMMARY:Physics Informed Machine Learning
LOCATION: Кочновский проезд, дом 3, лекционный зал Декарт на 3 этаже
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:Physics Informed Machine Learning\n\nMichael (Misha) Chertkov, Skoltech - A
 djunct Professor, on leave from Los Alamos National Laboratory)\n\nMachine 
 Learning (statistical engineering) capabilities are in a phase of tremendou
 s growth. Underlying these advances is a strong and deep connection to vari
 ous aspects of statistical physics. There is also a great opportunity in po
 inting these tools toward physical modeling. In this colloquium I illustrat
 e the two-way flow of ideas between physics and statistical engineering on 
 three examples. First, I review the work on structure learning and statisti
 cal estimation in power system distribution (thus physical) networks. Then 
 I describe recent progress in constructive understanding of graph learning 
 (on example of inverse Ising model) illustrating that the generic inverse t
 ask (of learning) is computationally easy in spite of the fact that the dir
 ect problem (inference or sampling) is difficult. I conclude speculating ho
 w macro-scale models of physics (e.g. large eddy simulations of turbulence)
  can be learned from micro-scale simulations (e.g. of Navier-Stocks equatio
 ns).\n\nhttps://cs.hse.ru/announcements/179173208.html\n\nРегистрация\nhttp
 s://docs.google.com/forms/d/1Z0SJ2hWn4zbzga0yYdX23RWTguQ-eIxzN8dQNKAZTZE/vi
 ewform This event was imported from: https://rtportal.ru/index.php/kalendar
 -sobytij/eventdetail/212/-/physics-informed-machine-learning?tmpl=component
DTSTAMP:20260812T205022Z
DTSTART;TZID=Europe/Moscow:20160412T181000
DTEND;TZID=Europe/Moscow:20160412T194000
SEQUENCE:0
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR