Traženje podgrupa iz višepoglednih podataka

master's thesis

• Title (english): Search for subgroups from multi-view data
• Author: Mislav Martinić
• Mentor: Matej Mihelčić (mentor)
• Committee member: Matej Mihelčić (predsjednik povjerenstva)
• Committee member: Mladen Vuković (član povjerenstva)
• Committee member: Goran Igaly (član povjerenstva)
• Committee member: Goranka Nogo (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Mathematics); Zagreb
• Defense date and country: 2021-12-02, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Mathematics

Abstract

U ovom diplomskom radu proučavamo CN2SD algoritam za traženje podgrupa te nudimo njegovo proširenje tako da traži podgrupe iz višepoglednih podataka. U uvodnom dijelu objašnjavamo što je traženje podgrupa, za što se koristi te gdje leži njezin značaj u dubinskoj analizi podataka. Predstavljamo CN2SD algoritam za traženje podgrupa te zrakasto pretraživanje, odnosno algoritam pretraživanja koji CN2SD koristi. Također predstavljamo pseudokod CN2SD algoritma, kao i opis njegovog rada. Glavni dio rada čine 2 prijedloga proširenja CN2SD algoritma za traženje podgrupa iz višepoglednih podataka, prijedlog nove ocjene dobrote podgrupa te pseudokodovi tih proširenja uz njihovo objašnjenje. Između ostalog radimo analizu vremenske složenosti CN2SD te oba proširenja. Na kraju smo proveli eksperiment na tri višepogledna skupa podataka, prikazali distribuciju podgrupa, usporedili obrađene pristupe traženju podgrupa te vremena izvršavanja algoritama.

Abstract (english)

In this graduate thesis we study CN2SD algorithm for subgroup discovery and propose its expenditure so it can be used for subgroup discovery from multiview data sets. In the introduction of this graduate thesis we explain what is subgroup discovery, what is it used for and where lies its importance in deep learning. We present CN2SD algorithm for subgroup discovery and Beam search, a search algorithm that CN2SD uses. We also present CN2SD pseudocode, as well as explanation of CN2SD. Main part of the thesis consists of two expanditure proposals of CN2SD algorithm for subgroup discovery from multiview data sets, proposal of new evaluation for subgroups induced from multiview data sets and pseudocodes of the two proposals with their explanation. Among other things we do time complexity analysis of CN2SD along with time complexity analysis of two expanditure proposals. In the end we conducted an experiment on three multiview data sets, showed subgroup distribution, compared elaborated approaches to subgroup discovery with corresponding algorithm execution times.

• Keywords: CN2SD algoritam; zrakasto pretraživanje; analiza vremenske složenosti
• Keywords (english): CN2SD algorithm; beam search; temporal complexity analysis
• Language: croatian
• URN:NBN: urn:nbn:hr:217:087279
• Study programme: Title: Computer Science and Mathematics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra računarstva i matematike (magistar/magistra računarstva i matematike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2021-12-15 11:26:04