Distinguished Professor: Mathematics, Tampere University in Finland.
Title: GUHA Reveals Secrets of Data
Abstract: In data mining, the aim is to find relations, associations, patterns, trends, and anomalies in data, here a flat matrix with rows and columns, whose cells can contain any symbols. GUHA is a particular data mining method introduced in 1960’s. GUHA is based on logic formalisms, not in statistics; associations found in the data are either true (supported by the data) or false (not supported by the data). The truth value of an association is based on contingency tables: there are millions of them based on the data. LISp-Miner is a software that takes the data as an input (preprocessed by the user), goes through it and outputs those that are true. We briefly present the mathematical foundations of GUHA logic, and then in detail a project that utilized GUHA data mining.
Biography:Professor Esko Turunen (born in 195 in Finland) is a distinguished mathematician affiliated with Tampere University in Finland. His academic journey includes earning a Doctor of Philosophy in Mathematics from Lappeenranta University of Technology in 1994 and a Licentiate of Philosophy in Mathematics from the University of Tampere in 1986. Before he got a chair in mathematics in Finland in 2014, we worked at Charles University in Prague and spent two years as a visiting researcher at Czech Academy of Sciences, and later another two years at the Technical University of Vienna as a Marie Curin senior fellow.
