On 25 February 2014, Opole University hosted Professor Jacek Koronacki, Head of the Institute of Computer Science of the Polish Academy of Sciences in Warsaw, who delivered his lecture under the title Statistical Inference for High-Dimensional Data: New Challenges brought by Life Sciences and other Research Areas. The meeting with Professor J. Koronacki, co-organized by the Institute of Mathematics and Informatics of Opole University, was held in Room 20 in the building of Collegium Civitas of Opole University at 89 Katowicka Street.
Professor Janusz Czelakowski welcomed the lecturer of the day and the gathered audience, and then introduced the problem area dealt with in the lecture. Next, Head of the Institute of Mathematics and Informatics, Professor Wiesłw Szwast, presented Professor J. Koronacki's scientific profile and output, and invited the lecturer to deliver his speech.
In his lecture, Professor J. Koronacki pointed to the development of biotechnology which makes it possible to collect a great deal of data concerning living cells, as well as to some challenges connected with it, including - among others - an analysis of observations of small sizes, each one of which can be characterized through a large number (the order of thousands) of attributes (features). As he writes: "major challenge in the analysis of many biological data matrices is due to their sizes: relatively small number of records (samples), often of the order of tens, versus thousands of attributes or features for each record. An obvious example are microarray gene expression experiments (here, the features are genes or, more precisely, their expression levels). Another, and a very specific one, is that of analyzing molecular interaction networks underlying HIV-1 resistance to reverse transcriptase inhibitors (here, the features are some physicochemical properties of amino acids)." The lecturer discussed questions connected with multiple testing of hypotheses and making use of different methods of prediction and selection of variables, including - among others - LASSO method and the method of random sub-space (RSM).
In the interesting discussion held after the lecture, various themes related to the subject presented in the lecture were raised and developed further.