Large-sample results for optimization-based clustering methods

Bryant, Peter G.

doi:10.1007/bf02616246

Cited by 38 publications

(27 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The careful monitoring of the maximum value attained by log-likelihoods like those in (2) and (3) while changing k has traditionally been applied as a method for choosing the number of clusters when α = 0. Moreover Bryant (1991) stated that the use of "weighted" log-likelihoods (3) is preferred to the use of log-likelihoods assuming equal weights (2). Notice that increasing k always causes Figure 8.…”

Section: Selecting the Number Of Groups And The Trimming Sizementioning

confidence: 99%

tclust: AnRPackage for a Trimming Approach to Cluster Analysis

Fritz¹,

García-Escudero²,

Mayo-Íscar³

2012

J. Stat. Soft.

View full text Add to dashboard Cite

Outlying data can heavily influence standard clustering methods. At the same time, clustering principles can be useful when robustifying statistical procedures. These two reasons motivate the development of feasible robust model-based clustering approaches. With this in mind, an R package for performing non-hierarchical robust clustering, called tclust, is presented here. Instead of trying to "fit" noisy data, a proportion α of the most outlying observations is trimmed. The tclust package efficiently handles different cluster scatter constraints. Graphical exploratory tools are also provided to help the user make sensible choices for the trimming proportion as well as the number of clusters to search for.

show abstract

Section: Selecting the Number Of Groups And The Trimming Sizementioning

confidence: 99%

tclust: AnRPackage for a Trimming Approach to Cluster Analysis

Fritz¹,

García-Escudero²,

Mayo-Íscar³

2012

J. Stat. Soft.

View full text Add to dashboard Cite

show abstract

“…The consideration of weights π j 's in classification likelihoods like in (2.2) goes back to Symons (1981). This criterium also appears in Bryant (1991) Since weights π j = 0 are possible, (2.2) does not necessarily increase strictly when increasing k. This fact was already noticed in Bryant (1991) in the untrimmed case α = 0. He also mentioned the possible merit of it in order to provide helpful guidance for choosing the number of groups in Clustering.…”

Section: The Tclust Methodologymentioning

confidence: 82%

Exploring the number of groups in robust model-based clustering

et al. 2010

View full text Add to dashboard Cite

Two key questions in Clustering problems are how to determine the number of groups properly and measure the strength of group-assignments. These questions are specially involved when the presence of certain fraction of outlying data is also expected.Any answer to these two key questions should depend on the assumed probabilisticmodel, the allowed group scatters and what we understand by noise. With this in mind, some exploratory "trimming-based" tools are presented in this work together with their justifications. The monitoring of optimal values reached when solving a robust clustering criteria and the use of some "discriminant" factors are the basis for these exploratory tools.

show abstract

“…We set g f ilt and g s to sufficiently small values to significantly reduce the number of variables after the filtering step and to allow the biological validation of the signature. For both datasets, the original partitions (1,2,3,4,5) and (1,2,3,4) show poor performances either according to their BIC ranking, to their ranking with the MCCV procedure (see figures III-B III-B) and to the generalization error rates associated to their respective signatures (minimum of gnagna and 38% respectively, data not shown).…”

Section: Classification Knowing Pmentioning

confidence: 96%

“…The choice of a mixture model allows to derive a Classification Log-Likelihood (CLL), as already proposed by [4]. Let P be the current partition.…”

Section: A Determination Of the Best Partitionmentioning

confidence: 99%

Grouping Levels of Exposure with Same Observable Effects before Class Prediction in Toxicogenomics

Guillemot

Philippe²,

Tenenhaus³

et al. 2008

2008 International Conference on Biocomputation, Bioinformatics, and Biomedical Technologies

View full text Add to dashboard Cite

Abstract-Gene expression profiling in toxicogenomics is often used to find molecular signature of toxicants. The range of doses chosen in toxicogenomics studies does not always represent all the possible effects on gene expression: several doses of toxicant can lead to the same observable effect on the transcriptome. This makes the problem of dose exposure prediction difficult to address. We propose a strategy allowing to gather the doses with similar effects prior to the computing of a molecular signature. The different gathering of doses are compared with criteria based on likelihood or Monte Carlo Cross Validation. The molecular signature is then determined via a voting algorithm. Experimental results point out that the obtained classifier has better prediction performances than the classifier computed according to the original labeling.

show abstract

Large-sample results for optimization-based clustering methods

Abstract: Classification, Clustering, Maximum likelihood, Asymptotic properties,

Cited by 38 publications

References 17 publications

tclust: AnRPackage for a Trimming Approach to Cluster Analysis

tclust: AnRPackage for a Trimming Approach to Cluster Analysis

Exploring the number of groups in robust model-based clustering

Grouping Levels of Exposure with Same Observable Effects before Class Prediction in Toxicogenomics

Contact Info

Product

Resources

About