Obtaining better quality final clustering by merging a collection of clusterings

Mimaroglu, Selim; Erdil, Ertunç

doi:10.1093/bioinformatics/btq489

Cited by 10 publications

(5 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This approach visualizes the base clustering solution as an undirected weighted graph, and Prim's algorithm is adapted to make a minimum-cost spanning tree of the weighted graph. Another approach proposed by the same authors also generates automatically the number of clusters, but requires to specify a relaxation parameter [14].…”

Section: Related Workmentioning

confidence: 99%

Multi-objective Consensus Clustering Framework for Flight Search Recommendation

Chatterjee¹,

Pasquier²,

Nanty³

et al. 2020

Preprint

View full text Add to dashboard Cite

In the travel industry, online customers book their travel itinerary according to several features, like cost and duration of the travel or the quality of amenities. To provide personalized recommendations for travel searches, an appropriate segmentation of customers is required. Clustering ensemble approaches were developed to overcome well-known problems of classical clustering approaches, that each rely on a different theoretical model and can thus identify in the data space only clusters corresponding to this model. Clustering ensemble approaches combine multiple clustering results, each from a different algorithmic configuration, for generating more robust consensus clusters corresponding to agreements between initial clusters. We present a new clustering ensemble multi-objective optimization-based framework developed for analyzing Amadeus customer search data and improve personalized recommendations. This framework optimizes diversity in the clustering ensemble search space and automatically determines an appropriate number of clusters without requiring user's input. Experimental results compare the efficiency of this approach with other existing approaches on Amadeus customer search data in terms of internal (Adjusted Rand Index) and external (Amadeus business metric) validations.

show abstract

Section: Related Workmentioning

confidence: 99%

Multi-objective Consensus Clustering Framework for Flight Search Recommendation

Chatterjee¹,

Pasquier²,

Nanty³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Since this algorithm is based on objects, it won't scale well. Selim Mimaroglu, Ertunc Erdi, 2010 [4] proposed Combining Multiple Clusterings Using Similarity graph (COMUSA), which is also graph based. This algorithm requires a relaxation parameter to find the number of true clusters.…”

Section: Related Workmentioning

confidence: 99%

Clustering fusion with automatic cluster number

Muneeswaran

Velvizhy

Kannan

2014

2014 International Conference on Recent Trends in Information Technology

View full text Add to dashboard Cite

Most of the real world applications use data clustering techniques for effective data analysis. All clustering techniques have some assumptions on the underlying dataset. We can get accurate clusters if the assumptions hold good. But it is difficult to satisfy all assumptions. Currently, not a single clustering algorithm is available to find all types of cluster shapes and structures. Therefore, an ensemble clustering algorithm is proposed in this paper in order to produce accurate clusters. Moreover, the existing clustering ensemble methods require more number of clusters in advance to produce final clusters. In this paper, we propose a novel method which groups a set of clusters into accurate final clusters to enhance the decision accuracy. This method does not need the number of clusters as input but produces the clusters automatically assuming the no of clusters. Index Terms-Clustering, Clustering ensemble.I.

show abstract

“…The similarity matrix is provided as input to an agglomerative clustering algorithm, as shown in Algorithm 1. [4] is a recent work for combining multiple clusterings into a final clustering, which is a graph-based method producing good results. COMUSA uses the evidence accumulated in the input clusterings, and produces a very good quality final clustering, where the number of clusters in the final clustering is obtained automatically with respect to relaxation rate.…”

Section: Combining Multiple Clusterings Using Evidence Accumulation (mentioning

confidence: 99%

“…Intracluster similarity concept can be expanded for a clustering i ðDÞ as shown in (4). Large values of ICS ð i ðDÞÞ express that clusters of i ðDÞ are compact, which is preferred.…”

Section: Finding the Best Clustering Automaticallymentioning

confidence: 99%

“…However, challenging characteristics of gene expression data sets such as high dimensionality and noise can easily degrade the output quality of even the most sophisticated clustering methods. Therefore, on a gene expression data set several clusterings can be generated and these clusterings can be supplied as input to a clustering ensemble method which produces better quality final clustering [3], [4]. Multiple input clusterings are generated on a data set by .…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

DICLENS: Divisive Clustering Ensemble with Automatic Cluster Number

Mimaroglu

Aksehirli

2012

IEEE/ACM Trans. Comput. Biol. and Bioinf.

View full text Add to dashboard Cite

Clustering has a long and rich history in a variety of scientific fields. Finding natural groupings of a data set is a hard task as attested by hundreds of clustering algorithms in the literature. Each clustering technique makes some assumptions about the underlying data set. If the assumptions hold, good clusterings can be expected. It is hard, in some cases impossible, to satisfy all the assumptions. Therefore, it is beneficial to apply different clustering methods on the same data set, or the same method with varying input parameters or both. We propose a novel method, DICLENS, which combines a set of clusterings into a final clustering having better overall quality. Our method produces the final clustering automatically and does not take any input parameters, a feature missing in many existing algorithms. Extensive experimental studies on real, artificial, and gene expression data sets demonstrate that DICLENS produces very good quality clusterings in a short amount of time. DICLENS implementation runs on standard personal computers by being scalable, and by consuming very little memory and CPU.

show abstract

Obtaining better quality final clustering by merging a collection of clusterings

Abstract: selim.mimaroglu@bahcesehir.edu.tr

Cited by 10 publications

References 2 publications

Multi-objective Consensus Clustering Framework for Flight Search Recommendation

Multi-objective Consensus Clustering Framework for Flight Search Recommendation

Clustering fusion with automatic cluster number

DICLENS: Divisive Clustering Ensemble with Automatic Cluster Number

Contact Info

Product

Resources

About