Prof. Milan Stehlík | October 18, 2019 | 15:00 | V.1.27
Abstract:
Extracting chaotical and stochastic parts of information from time series needs very specific techniques. Motivated by two applications, image processing for cancer discrimination and methane emissions modelling we will explain the necessary techniques for statistical learning on chaotical and stochastic parts from data. In particular, Tsallis Entropy will be introduced and its role in information theory for dynamical system explained. Iterated function systems will be used as an example for chaos re-simulation. Construction of stochastic fractals will be discussed. We will show the importance of decomposition of data to stochastic, deterministic and chaotic part.
CV:
Currently he is Visiting professor at the Department of Statistics & Actuarial Science, The University of Iowa. In 2018 he was visiting Full Professor at School of Mathematics & Statistical Sciences Arizona State University, AZ, USA. He was involved in several international projects and collaborations in Austria, Spain, Russia, Canada, Germany, USA among others.
He does research in Extremes, Optimal design of experiments, Statistical Modelling, Neural Computing, Cancer discrimination. He servers as Associate Editor for Europe of Neural Computing and Applications, Associate Editor of Journal of Applied Statistics and Revstat. He has been Principal Investigator of Innovative project LIT-2016-1-SEE-023 Title: Modeling complex dependencies: how to make strategic multicriterial decisions? at Linz Institute of Technology, Austria and Chilean FONDECYT Regular. He published more than 180 papers and gave more than 190 talks.
