Keynote Lecture

Learning data structures with inherent complex logic

Prof. Wlodzislaw Duch
Dept. of Informatics, Nicolaus Copernicus University
Torun, Poland
http://www.is.umk.pl/~duch/

Abstract: The greatest challenge for computational intelligence is to learn in difficult, highly non-separable situations. Current state-of-the-art learning algorithms are useful only when data is linearly separable using appropriate kernels. Even simple problems with non-trivial logic, like parity problems, cannot be learned with such algorithms. Many problems in bioinformatics and text analysis require complex logic or discovery of (approximate) logical structure in the data. Visualization of learning dynamics in neural networks shows that frequently separability cannot be achieved, but simpler goals for learning may be set. k-separability, or the projection of data on a line and segmentation into intervals, is an interesting concept that allows for estimation of the degree of non-separability. Difficult problems may be learned in this way although quite different algorithms are required.

Brief Biography of the Speaker:
Wlodzislaw Duch heads the Department of Informatics, Nicolaus Copernicus University, Torun, Poland, and is a Visting Professor at Nanyang Technological University, Singapore (2003-7). Ph.D. in quantum chemistry (1980), postdoc at USC, Los Angeles (1980-82), D.Sc. in applied math (1987); worked at University of Florida; Max-Planck-Institute, Munich, Germany, Kyushu Institute of Technology, Meiji and Rikkyo University in Japan, and several other institutions. He is on the editorial board of IEEE TNN, CPC, NIP-LR, Journal of Mind and Behavior, and 7 other journals; co-founder & scientific editor of the "Polish Cognitive Science" journal; president of the European Neural Networks Society (2006-2008), member of IEEE NNS Technical committee; expert of the European Union science programs; published over 350 scientific and popular articles, 4 books, edited many others, his DuchSoft company makes GhostMiner software package marketed by Fujitsu. Expert in computational intelligence (CI), especially methods that facilitate understanding of data, and algorithms inspired by models of brain functions at different levels. Among other topics s on creation of general CI theory based on similarity evaluation, meta-learning schemes that automatically discover the best model for a given data, geometrical theories for modeling of mental events and relating such models to neurodynamics, and toys that facilitate mental development. With a wide background in many branches of science and understanding of different cultures he bridges various scientific communities. As a service to the international community maintains many web pages related to CI, computational neuroscience, machine learning and statistics. To access these pages and his full CV type "Duch" in Google.