📜 ⬆️ ⬇️

AI @ MIPT: “Neuromorphic computations and brain mechanisms”

image

February 12 at Fiztekh will take place the next meeting from the series of seminars on artificial intelligence “Neuromorphic calculations and brain mechanisms”. Vitaly Dunin-Barkovsky , head of the department of neuroinformatics at the Center for Optical Neural Technologies and principal investigator of the iPavlov project, will give examples on the general features of the “intelligence” of neuromorphic structures and the nervous system.

The seminar will begin at 18:30 in the audience of 107 MIPT Biokorpus. For full-time participation, you must first register . The lecture will also be broadcast online on the official page of the Teacher of Applied Science and Technology on VKontakte .

The success of multilayered neuromorphic structures for a long time was ensured solely by the mechanism of back propagation of errors. This mechanism is unrealizable in living neural systems, so scientists continued to search for secrets of brain efficiency. Recent computational results have shown that schemes that allow physiological implementation can be used to train or tune multilayer neural networks. The “intelligence” of neuromorphic structures and the nervous system can have a common source. The report on specific examples traces the consequences of such a generality for both neuroscience and neuromorphic structures. This serves as a basis for discussing the prospects and methods for creating general intelligence systems.
')
Biography of the lecturer
Vitaly Dunin-Barkovsky - Doctor of Physical and Mathematical Sciences, Professor. He graduated from the Faculty of Radio Engineering and Cybernetics of the Moscow Institute of Physics and Technology, then worked at the Institute for Information Transmission Problems of the Russian Academy of Sciences and Moscow Technological University (MIREA), and also was the director of the Institute of Neurocybernetics Research Institute A. B. Kogan Rostov State University. From 1997 to 2006 he was a visiting professor at Texas Tech University. The first editor of Frontiersin in the direction of Towards and Beyond is the Human Level of Artificial Intelligence and a member of the editorial boards of such journals as Neurocomputing, Optical Memory and Neural Networks, as well as Neurocomputers and their applications. Holds weekly working sessions “Neuronet Cauldron of Ideas” at the Agency for Strategic Initiatives.

The course materials AI @ MIPT are available in the Telegram-channel and Vkontakte group.

Source: https://habr.com/ru/post/348330/


All Articles