[Seminar] "Time-universal data compression and prediction” by Prof. Boris Ryabko

*** Time/location has been changed to 10:30-11:30am@ Lab3 C700 ****

Speaker: Prof. Boris Ryabko

Head of Laboratory of data protection, ICT, Siberian Branch of RAS.

More infomration can be found at https://boris.ryabko.net

[Abstract]

We describe a new approach to building data compressors and predictors for time series. The proposed approach allows combining the set of methods into one so that the resulting method is as effective as the best of the set, and the required additional calculation is very small.

Note that there are currently many artificial intelligence methods that are used inside data compressors and predictors. Thus, the proposed approach makes it possible to use an AI method that is most suitable for specific data.

Host: Jonathan Miller, Professor, Physics and Biology Unit.

Prof. Ryabko will be visiting OIST from September 16th to September 20th.  If you would like to meet with Prof. Ryabko during his visit, please contact Midori Tanahara <e-mail: midori.tanahara [at] oist.jp>.