Implementation and evaluation of RIPPER algorithm for constructing covering rules

U ovom radu je opisan RIPPERk algoritam i njegov prethodnik IREP te najvažniji principi po kojima rade. Također je opisana teorijska podloga za rad alogirtama strojnog učenja. Nakon teorijskog dijela opisana je implementacija RIPPERk algoritma te su napisane upute za instalaciju. Zatim je opisan način korištenja sličnih algoritama s javno dostupnih izvora, te su na kraju prikazani i uspoređeni rezultati različitih implementacija.This paper describes the RIPPERk algorithm , its predecessor IREP and the most important principles by which they work. Theoretical background for the work of machine learning algorithms is also described. After the theoretical section, the implementation of the RIPPERk algorithm is described and the installation instructions are written. Next, the usage of similar algorithms which are found in publicly available sources is explained. Finally the results of the various implementations are presented and compared