Celovit pregled orodij za samodejno strojno učenje

diplomsko delo

Tomi Milošič (Author), Iztok Fister (Mentor), Špela Pečnik (Co-mentor)

Abstract

V diplomski nalogi smo raziskali področje samodejnega strojnega učenja, osredotočili smo se na orodja za samodejno strojno učenje in poudarili njihove prednosti in slabosti na podlagi primerjave glede na različne nabore podatkov. Osredotočili smo se tudi na metodo klasifikacije, saj je to pogosta naloga strojnega učenja. Namen diplomske naloge je ugotoviti, katero orodje je najbolj optimalno za posamezno nalogo. Diplomsko nalogo smo razdelili na dva dela, in sicer teoretični del in praktični del. V teoretičnem delu smo se osredotočili na razjasnitev pojmov, zgodovino strojnega učenja in opis orodij samodejnega strojnega učenja. V praktičnem delu smo opravili primerjave med orodji in ugotovili, da le-ta vračajo podobne rezultate klasifikacije različno hitro. Ugotovili smo tudi, da so orodja namenjena uporabnikom, ki niso strokovnjaki na področju strojnega učenja, in da si orodja delijo skupne značilnosti.

Keywords

klasifikacija;strojno učenje;samodejno strojno učenje;umetna inteligenca;Na spremnem listu tudi nasl. v angl.:;

Data

Language:	Slovenian
Year of publishing:	2021
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[T. Milošič]
UDC:	004.85(043.2)
COBISS:	94366979
Views:	285
Downloads:	50
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	A comprehensive overview of automated machine learning tools
Secondary abstract:	In the thesis, we researched the field of automatic machine learning, focused on automatic machine learning tools, and highlighted their advantages and disadvantages based on comparison with different data sets. The focus was also on classification which is a common task of machine learning. The purpose of the thesis was to determine which tool is most optimal for individual tasks. The thesis was divided into two parts, a theoretical part, and an empirical part. In the theoretical part, the clarification of concepts, the history of machine learning, and the description of automatic machine learning tools was presented. In the empirical part, comparisons were made between tools and it was found that AutoML tools return similar classification results at different speeds. We also found that the tools are intended for users who are not experts in the field of machine learning and that the tools share common features.
Secondary keywords:	classification;machine learning;automated machine learning;artificial intelligence;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Informatika in tehnologije komuniciranja
Pages:	XII, 52 str.
ID:	13285157