Cluster analysis and the identification of aggregations

Strauss, Richard E.

doi:10.1006/anbe.2000.1608

Cited by 11 publications

(8 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A second problem is that the minimization of this criterion has the tendency to yield equally sized clusters, even when the actual classes in the data contain different numbers of objects. The tendency of the K-means cluster analysis to produce spherical and equally sized clusters has been documented by many authors (Eldershaw & Hegland, 1997;Everitt, 1993;Gordon, 1999;Jain & Dubes, 1988;Scott & Symons, 1971;Strauss, 2001). However, the authors do not describe the magnitude of these effects on the recovery of clusters.…”

mentioning

confidence: 88%

Effects of Group Size and Lack of Sphericity on the Recovery of Clusters in K-means Cluster Analysis

Craen¹,

Commandeur

Frank³

et al. 2006

Multivariate Behavioral Res.

View full text Add to dashboard Cite

K-means cluster analysis is known for its tendency to produce spherical and equally sized clusters. To assess the magnitude of these effects, a simulation study was conducted, in which populations were created with varying departures from sphericity and group sizes. An analysis of the recovery of clusters in the samples taken from these populations showed a significant effect of lack of sphericity and group size. This effect was, however, not as large as expected, with still a recovery index of more than 0.5 in the "worst case scenario." An interaction effect between the two data aspects was also found. The decreasing trend in the recovery of clusters for increasing departures from sphericity is different for equal and unequal group sizes.

show abstract

mentioning

confidence: 88%

Effects of Group Size and Lack of Sphericity on the Recovery of Clusters in K-means Cluster Analysis

Craen¹,

Commandeur

Frank³

et al. 2006

Multivariate Behavioral Res.

View full text Add to dashboard Cite

show abstract

“…To estimate the number and compositions of groups of species in the MDS coordinate space, we used a "k nearest-neighbor" (kNN) classification procedure (Strauss, 2001). We modified the nearest-neighbor algorithm described by Knorr et al (2000) to identify and omit outliers from the recognized groups based on k rather than all nearest neighbors.…”

Section: Animal Burrow or Hole (Abh)mentioning

confidence: 99%

Roosting Ecology of Amazonian Bats: Evidence for Guild Structure in Hyperdiverse Mammalian Communities

Voss

Fleck

Strauss

et al. 2016

American Museum Novitates

Self Cite

View full text Add to dashboard Cite

The ecological mechanisms that sustain high species richness in Neotropical bat communities have attracted research attention for several decades. Although many ecologists have studied the feeding behavior and diets of Neotropical bats on the assumption that food is a limiting resource, other resource axes that might be important for species coexistence are often ignored. Diurnal refugia, in particular, are a crucial resource for bats, many of which exhibit conspicuous morphological or behavioral adaptations to the roost environment. Here we report and analyze information about roost occupancy based on >500 field observations of Amazonian bats. Statistical analyses of these data suggest the existence of distinct groups of species roosting (1) in foliage, (2) exposed on the trunks of standing trees, (3) in cavities in standing trees, (4) in or under fallen trees, (5) beneath undercut earth banks, and (6) in arboreal insect nests; additionally, we recognize other groups that roost (7) in animal burrows, and (8) in rocks or caves. Roosting-guild membership is hypothesized to have a filtering effect on Amazonian bat community composition because some types of roosts are absent or uncommon in certain habitats. Among other applications of our results, cross-classifying bat species by trophic and roosting guilds suggests that the often-reported deficit of gleaning animalivores in secondary vegetation by comparison with primary forest might reflect habitat differences in roost availability rather than food resources. In general, ecological and evolutionary studies of Neotropical bats would be enhanced by considering both trophic-and roosting-guild membership in future analyses, but additional fieldwork will be required to determine the roosting behavior of many data-deficient species.

show abstract

“…There exist a wide variety of clustering and classification techniques, such as hierarchical cluster analysis, nearest neighbor classification, and techniques based on swarm intelligence algorithms (e.g., [15,16]). An iterative, kth nearest neighbor, hierarchical cluster analysis for detecting shoals was been proposed in [6]. However, results from hierarchical cluster analysis require a cutoff criteria in order to determine the actual clusters.…”

Section: Affinity Propagationmentioning

confidence: 99%

“…According to [6], objectivity of the clustering results can be ensured by comparing them with how human observers partition the same data. Thus, by comparing to human observers, we obtain a type of relative validation in which human performance is considered as if it were an alternative clustering method.…”

Section: Comparing Affinity Propagation Clustering With Human Clusteringmentioning

confidence: 99%

“…Regarding the analysis of animal aggregations, it is accepted that determining objectively the number of groups present and which individuals are members of each group is a very difficult task [6]; consequently, a quantitative method is needed to detect multiple groups. Moreover, in order to determine the validity of such a method, its outcome should be compared to estimates made by human observers, which would "ensure consistency and objectivity across time and among different observers and studies" [6, p. 483].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Determining shoal membership using affinity propagation

Quera

Beltran

Givoni

et al. 2013

Behavioural Brain Research

View full text Add to dashboard Cite

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues.Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. We propose using the affinity propagation clustering for detecting multiple shoals. A soft temporal constraint is included in order to detect shoal fusion and fission. We explore how affinity propagation performs on agent-based simulated shoals. We compare affinity propagation clustering to human clustering of the same data. Affinity propagation is an appealing approach for detecting shoal dynamics. a r t i c l e i n f o b s t r a c tWe propose using the affinity propagation (AP) clustering algorithm for detecting multiple disjoint shoals, and we present an extension of AP, denoted by STAP, that can be applied to shoals that fusion and fission across time. STAP incorporates into AP a soft temporal constraint that takes cluster dynamics into account, encouraging partitions obtained at successive time steps to be consistent with each other. We explore how STAP performs under different settings of its parameters (strength of the temporal constraint, preferences, and distance metric) by applying the algorithm to simulated sequences of collective coordinated motion. We study the validity of STAP by comparing its results to partitioning of the same data obtained from human observers in a controlled experiment. We observe that, under specific circumstances, AP yields partitions that agree quite closely with the ones made by human observers. We conclude that using the STAP algorithm with appropriate parameter settings is an appealing approach for detecting shoal fusion-fission dynamics.

show abstract

Cluster analysis and the identification of aggregations

Cited by 11 publications

References 54 publications

Effects of Group Size and Lack of Sphericity on the Recovery of Clusters in K-means Cluster Analysis

Effects of Group Size and Lack of Sphericity on the Recovery of Clusters in K-means Cluster Analysis

Roosting Ecology of Amazonian Bats: Evidence for Guild Structure in Hyperdiverse Mammalian Communities

Determining shoal membership using affinity propagation

Contact Info

Product

Resources

About