Razvoj vtičnika za podporo strojnemu učenju v programu Jamovi

diplomsko delo

Maruša Konečnik (Author), Sašo Karakatič (Mentor)

Abstract

Glavni cilj diplomskega dela je bil razvoj vtičnika za program za statistično analizo Jamovi. Poleg tega smo v diplomskem delu predstavili osnove strojnega učenja, z njim povezane metrike ocenjevanja ter opisali in primerjali nekaj programov za statistično analizo. Vtičnik omogoča podporo strojnemu učenju za različne algoritme. Osredotočili smo se na algoritma odločitvenega drevesa enojno odločitveno drevo ter naključni gozd Omogočena je tudi primerjava rezultatov več algoritmov. V praktičnem delu diplomske naloge smo razvili delujoč vtičnik, v teoretičnem delu pa smo zapisali osnovna navodila za razvoj vtičnika za Jamovi ter opisali funkcionalnosti ter uporabo nastalega vtičnika.

Keywords

strojno učenje;statistična analiza;Jamovi;metrike ocenjevanja;vtičniki;diplomske naloge;

Data

Language:	Slovenian
Year of publishing:	2020
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[M. Konečnik]
UDC:	004.42:004.85(043.2)
COBISS:	37496067
Views:	517
Downloads:	63
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Development of machine learning plugin for Jamovi
Secondary abstract:	Our main objective was to develop a machine learning plugin for Jamovi. Besides that, we presented the basics of machine learning and evaluation metrics. We have described and compared a few statistical analysis programs. While developing the plugin we have focused on single decision tree and random forest algorithms. There is also possible to compare results and performance of more algorithms at the same time. In the practical part of our thesis we have developed a working plugin, theoretical part describes a basic manual for developing a plugin for Jamovi and also describes the work and the functionalities of our own plugin.
Secondary keywords:	machine learning;statistical analysis;Jamovi;evaluation metrics;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Informatika in tehnologije komuniciranja
Pages:	IX, 49 f.
ID:	11889547