On the Impact of Dissimilarity Measure in k-Modes Clustering Algorithm

Ng, Michael K.; Li, M.J.; Huang, Jiaoying; He, Zengyou

doi:10.1109/tpami.2007.53

Cited by 187 publications

(108 citation statements)

References 8 publications

Supporting

Mentioning

102

Contrasting

Order By: Relevance

“…The algorithm proposed in this paper is different from k-Median algorithm in reference [2], this paper uses k-Modes clustering algorithm [3] to label unlabeled data in the process of incrementally creating decision tree. Because k-Modes clustering algorithm is suitable for dealing with discrete attributes, we use discretization method [4] for dealing with continuous attributes.…”

Section: Semi-supervised Learning Methods Based On Kmodes Algorithm Anmentioning

confidence: 99%

Coal Mine Safety Evaluation Method Based on Incomplete Labeled Data Stream Classification

Sun¹,

Zhou²,

Sun³

2014

TOCSJ

View full text Add to dashboard Cite

Monitoring data in coal mine is essentially data stream, and missing coal mine monitoring data is caused by harsh coal mine environment, therefore coal mine safety evaluation can be seen as incomplete labeled data stream classification. The method is proposed for unlabeled data and concept drift in incomplete labeled data stream in this paper that uses semi-supervised learning method based on k-Modes algorithm and incremental decision tree model and concept drift detection mechanism based on clustering concept-cluster. Experimental results show the method can better label unlabeled data and detect concept drift in incomplete labeled data stream, and it has better classification accuracy for incomplete labeled data stream, and it provides a new practical approach for coal mine safety evaluation.

show abstract

Section: Semi-supervised Learning Methods Based On Kmodes Algorithm Anmentioning

confidence: 99%

Coal Mine Safety Evaluation Method Based on Incomplete Labeled Data Stream Classification

Sun¹,

Zhou²,

Sun³

2014

TOCSJ

View full text Add to dashboard Cite

show abstract

“…Dissimilarity based k-mode [18] is also the one of the K-Mode Algorithm in which a new dissimilarity measure is proposed in which the modes of clusters were updated in each iteration ad utilizes some theorems to update the mode of the cluster.…”

Section: Literature Surveymentioning

confidence: 99%

Comparative Study of Initial Centroid Based K-Mode Algorithm

2017

IJAERD

View full text Add to dashboard Cite

“…The smaller the number of mismatches, the more similar are two objects. This measure is also a kind of generalized Hamming distance (Ng et al, 2007).…”

Section: Definitions and Notationsmentioning

confidence: 99%

Hierarchical Clustering with Simple Matching and Joint Entropy Dissimilarity Measure

Ergüt

2014

J. Mod. App. Stat. Meth.

View full text Add to dashboard Cite

Conventional clustering algorithms are restricted for use with data containing ratio or interval scale variables; hence, distances are used. As social studies require merely categorical data, the literature is enriched with more complicated clustering techniques and algorithms of categorical data. These techniques are based on similarity or dissimilarity matrices. The algorithms are using density based or pattern based approaches. A probabilistic nature to similarity structure is proposed. The entropy dissimilarity measure has comparable results with simple matching dissimilarity at hierarchical clustering. It overcomes dimension increase through binarization of the categorical data. This approach is also functional with the clustering methods, where apriori cluster number information is available.

show abstract

On the Impact of Dissimilarity Measure in k-Modes Clustering Algorithm

Cited by 187 publications

References 8 publications

Coal Mine Safety Evaluation Method Based on Incomplete Labeled Data Stream Classification

Coal Mine Safety Evaluation Method Based on Incomplete Labeled Data Stream Classification

Comparative Study of Initial Centroid Based K-Mode Algorithm

Hierarchical Clustering with Simple Matching and Joint Entropy Dissimilarity Measure

Contact Info

Product

Resources

About