Novel approaches of data-mining in experimental physics

Abstract

Data mining for processing experimental data in high energy and
nuclear physics led to many multiparametric problems, two of them are consid-
ered: (i) hypothesis testing and classication approaches based on articial neural
networks and boosted decision trees (ii) clustering of large amounts of data by
so-called growing neural gas. Some examples from the practice of the Joint In-
stitute for Nuclear research are given to show how to prepare data to deal with
those approaches.