Constrained Intelligent K-Means: Improving Results with Limited Previous Knowledge

Renato Cordeiro de Amorim

doi:10.1109/ADVCOMP.2008.30

Constrained Intelligent K-Means: Improving Results with Limited Previous Knowledge

Renato Cordeiro de Amorim

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

13 Citations (Scopus)

165 Downloads (Pure)

Abstract

It is here presented a new method for clustering that uses very limited amount of labeled data, employees two pairwise rules, namely must link and cannot link and a single wise one, cannot cluster. It is demonstrated that the incorporation of these rules in the intelligent k-means algorithm may increase the accuracy of results, this is proven with experiments where the real number of clusters in the data is unknown to the method

Original language	English
Title of host publication	Procs of the Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008
Subtitle of host publication	ADVCOMP'08
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	176-180
Number of pages	5
ISBN (Electronic)	978-0-7695-3369-8
ISBN (Print)	978-0-7695-3369-8
DOIs	https://doi.org/10.1109/ADVCOMP.2008.30
Publication status	Published - 2008
Event	Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008. ADVCOMP'08 - Valencia, Spain Duration: 29 Sept 2008 → 4 Oct 2008

Conference

Conference	Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008. ADVCOMP'08
Country/Territory	Spain
City	Valencia
Period	29/09/08 → 4/10/08

Access to Document

10.1109/ADVCOMP.2008.30

Amorim-ClusteringSubmitted manuscript, 388 KB

Cite this

@inproceedings{5cde0a1832384c6ea87407ea81b3d32a,

title = "Constrained Intelligent K-Means: Improving Results with Limited Previous Knowledge",

abstract = "It is here presented a new method for clustering that uses very limited amount of labeled data, employees two pairwise rules, namely must link and cannot link and a single wise one, cannot cluster. It is demonstrated that the incorporation of these rules in the intelligent k-means algorithm may increase the accuracy of results, this is proven with experiments where the real number of clusters in the data is unknown to the method",

author = "{de Amorim}, {Renato Cordeiro}",

year = "2008",

doi = "10.1109/ADVCOMP.2008.30",

language = "English",

isbn = "978-0-7695-3369-8",

pages = "176--180",

booktitle = "Procs of the Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

address = "United States",

note = "Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008. ADVCOMP'08 ; Conference date: 29-09-2008 Through 04-10-2008",

}

de Amorim, RC 2008, Constrained Intelligent K-Means: Improving Results with Limited Previous Knowledge. in Procs of the Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008: ADVCOMP'08. Institute of Electrical and Electronics Engineers (IEEE), pp. 176-180, Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008. ADVCOMP'08, Valencia, Spain, 29/09/08. https://doi.org/10.1109/ADVCOMP.2008.30

Constrained Intelligent K-Means: Improving Results with Limited Previous Knowledge. / de Amorim, Renato Cordeiro.
Procs of the Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008: ADVCOMP'08. Institute of Electrical and Electronics Engineers (IEEE), 2008. p. 176-180.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Constrained Intelligent K-Means

T2 - Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008. ADVCOMP'08

AU - de Amorim, Renato Cordeiro

PY - 2008

Y1 - 2008

N2 - It is here presented a new method for clustering that uses very limited amount of labeled data, employees two pairwise rules, namely must link and cannot link and a single wise one, cannot cluster. It is demonstrated that the incorporation of these rules in the intelligent k-means algorithm may increase the accuracy of results, this is proven with experiments where the real number of clusters in the data is unknown to the method

AB - It is here presented a new method for clustering that uses very limited amount of labeled data, employees two pairwise rules, namely must link and cannot link and a single wise one, cannot cluster. It is demonstrated that the incorporation of these rules in the intelligent k-means algorithm may increase the accuracy of results, this is proven with experiments where the real number of clusters in the data is unknown to the method

U2 - 10.1109/ADVCOMP.2008.30

DO - 10.1109/ADVCOMP.2008.30

M3 - Conference contribution

SN - 978-0-7695-3369-8

SP - 176

EP - 180

BT - Procs of the Second Int Conf on Advanced Engineering Computing and Applications in Sciences, 2008

PB - Institute of Electrical and Electronics Engineers (IEEE)

Y2 - 29 September 2008 through 4 October 2008

ER -

Constrained Intelligent K-Means: Improving Results with Limited Previous Knowledge

Abstract

Conference

Access to Document

Fingerprint

Cite this