Please use this identifier to cite or link to this item: http://hdl.handle.net/10316/48034
Title: Um breve estudo sobre o algoritmo K-means
Authors: Nunes, Diogo Henriques Freitas 
Orientador: Santos, José Luis Esteves dos
Keywords: Clustering; Clusters; K-means; Expectation-Maximization; Kernel K-means
Issue Date: 5-Jul-2016
Abstract: Este trabalho teve como objetivo aprofundar o conhecimento de uma ferramenta de data mining conhecida como o clustering. Optou-se por direcionar este estudo para um algoritmo de clustering clássico, o K-means, e prova-se a sua convergência. Complementarmente apresentam-se outros dois algoritmos, o Expetation-Maximization e o Kernel K-means, e realizam-se testes de desempenho entre os três. No final aplicaram-se estes algoritmos a vários tipos de problemas nomeadamente no campo da segmentação de imagens.
The objective of this work was to understand a data mining technique known as clustering. Particulary, this work was focused on a classic algorithm known as K-means and demonstrating its convergence. To understand its potential two more classic algorithms were studied, named Expetation-Maximization and Kernel K-means and their respective performances were tested and compared. Finally, these algorithms were applied on image segmentation.
URI: http://hdl.handle.net/10316/48034
Rights: openAccess
Appears in Collections:FCTUC Matemática - Teses de Mestrado

Files in This Item:
File Description SizeFormat
Tese_DiogoNunes.pdf1.96 MBAdobe PDFView/Open
Show full item record

Page view(s)

116
checked on Sep 16, 2019

Download(s) 10

1,483
checked on Sep 16, 2019

Google ScholarTM

Check


This item is licensed under a Creative Commons License Creative Commons