Implementacija odločitvenih dreves v programskem jeziku Python

diplomsko delo

Gregor Gorjanc (Author), Vili Podgorelec (Mentor)

Abstract

V diplomskem delu je predstavljeno področje strojnega učenja, del katerega so odločitvena drevesa. Čeprav je odločitveno drevo v osnovi pripomoček za vizualizacijo, nas v sklopu strojnega učenja zanima proces gradnje. Obstaja več algoritmov, ki s pomočjo podatkovnih množic generirajo odločitvena drevesa. V delu je podrobno predstavljeno delovanje algoritma C4.5, čigar implementacija predstavlja jedro naloge. Algoritem je bil implementiran s pomočjo programskega jezika Python. Po osnovni implementacijo je bil izveden proces optimizacije, kjer so preizkušene različne strukture programskega jezika Python. Po optimizaciji je bila izvedena primerjalna analiza delovanja. Izvedena je bila tudi primerjava delovanja implementiranega algoritma z obstoječo implementacijo J48.

Keywords

strojno učenje;odločitveno drevo;implementacija;programski jezik Python;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:	[G. Gorjanc]
UDC:	004.85.021:004.43(043.2)
COBISS:	27066115
Views:	1083
Downloads:	224
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Implementation of decision trees in the Python programming language
Secondary abstract:	In this diploma thesis we researched the field of machine learning, more specifically the decision trees. There are several different algorithms for building the decision trees. One of those algorithms is C4.5, whose implementation represents the main part of this diploma thesis. The algorithm was implemented using the Python programming language. After the completed implementation, we carried out the process of optimization, during which we tested different structures of the Python programming language. Following the optimization was the comparison of different stages of the optimization. As a part of analysis, we also compared the implementation to the existing implementation J48.
Secondary keywords:	machine learning;decision trees;Python;C4.5;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Računalništvo in informacijske tehnologije
Pages:	IX, 47 str.
ID:	11443071