BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//jEvents 2.0 for Joomla//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:cor3id1j60rmab9h65hjcb9k6hhj0b9ocgp36bb171hm2dr168ojgp9hck@google.com
CATEGORIES:RTportal.ru
CREATED:20170606T092329
SUMMARY:семинар "Структурные модели и глубинное обучение": дополнительный семинар
LOCATION:ИППИ РАН, аудитория 615
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:Коллеги!\n\nНа дополнительном заседании семинара "Структурные модели и глуб
 инное обучение", которое состоится 17 июня (пятница), в 16.00, ИППИ РАН (ht
 tp://iitp.ru/ru/contacts.htm), 6 этаж, 615 аудитория\n\nбудет представлено 
 несколько коротких сообщений о различных интересных направлениях машинного 
 обучения.\n\n===================\n\nAuthor: Denis Volhonskiy (HSE, IITP)\n\
 nTitle: Deep Convolutional Generative Adversarial Networks in Steganography
 \n\nAnnotation: Steganography is the way of hiding information within other
  information (called container). Steganalysis is the study of detecting mes
 sages hidden using steganography --- usually it is binary classifier (detec
 t if there is some information / no information in container).\n\nWe propos
 e a new model based on Deep Convolutional Generative Adversarial Networks f
 or generating images, that could be used for more safety information embedd
 ing (in terms of steganalysis accuracy) using steganography algorithms. Our
  model allows increasing steganalysis error on a test set of generated imag
 es in comparison with an original test set.\n\n===================\n\nAutho
 r: Vladislav Ishimtsev (HSE, IITP)\n\nTitle: Conformalized density- and dis
 tance-based anomaly detection in time-series data\n\nAnnotation: Most of th
 e world's data is streaming, time series data, where anomalies give pertine
 nt information in critical situations; Examples abound in such fields as fi
 nance, IT, security, health and energy. However, detection of anomalies in 
 time-series data is a difficult task, requiring detectors to process data i
 n real-time, not batches, and learn while simultaneously making predictions
 . \n\nWe consider new approaches to detect anomalies in time-series data us
 ing conformalized density- and distance-based anomaly detection algorithms.
  \n\nTesting and comparison of algorithms will be done on Numenta Anomaly B
 enchmark (NAB), which attempts to provide a controlled and repeatable envir
 onment of open-source tools to test and measure anomaly detection algorithm
 s on streaming data.\n\n===================\n\nAuthor: Albert Matveev (HSE)
 \n\nTitle: Experts Aggregation for Time Series Prediction\n\nAnnotation: Ti
 me series prediction is one of the most significant problems in applied mat
 hematics. Estimation of a future outcome of some sequence is desired in man
 y scientific and practical fields, especially in finance. Regression-type f
 orecasters are usually used in order to solve this problem. However, it is 
 essential to note that for different regimes we need to use different metho
 ds. It is obvious that without knowing true future value of a time series, 
 the learner cannot select a method and corresponding forecast, which would 
 provide the best accuracy among available ones. Therefore, we consider a se
 t of independent base forecasting algorithms which we call experts and form
 ulate a problem of prediction with expert advice. We would like to construc
 t an aggregating algorithm that will adaptively aggregate available expert 
 forecasts so that aggregated forecast is close to the best expert in terms 
 of loss process. \n\nIn this talk the main aggregating schemes and loss bou
 nds for them will be presented. These aggregating algorithms will be compar
 ed on several datasets and the results of computational experiments will be
  provided.\n\n===================\n\nAuthor: Oleg Maslennikov (HSE)\n\nTitl
 e: Deep Features based Image Retrieval\n\nAnnotation: Computer vision tasks
  are common in today's world, for example, OCR tasks (optical character rec
 ognition), license plate recognition with traffic cameras, image processing
  in medicine and others. One of these tasks is CBIR - content base image re
 trieval. Examples of such tasks are the image search in Google, recently ap
 peared service “Findface”, which allows finding a person in social networks
  by his/her photo.\n\nIn the presentation we will review approaches to deep
  learning based solution of CBIR tasks, discuss various architectures of co
 rresponding neural networks, as well as the choice of a metric distance and
  its influence on obtained results.\n\n--\n\nС уважением,\n\nЕвгений Бурнае
 в This event was imported from: https://rtportal.ru/index.php/kalendar-soby
 tij/eventdetail/136/-/seminar-strukturnye-modeli-i-glubinnoe-obuchenie-dopo
 lnitelnyj-seminar?tmpl=component
DTSTAMP:20260812T210332Z
DTSTART;TZID=Europe/Moscow:20160617T160000
DTEND;TZID=Europe/Moscow:20160617T173000
SEQUENCE:0
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR