A dendrite method for cluster analysis

Caliński, Tadeusz; Harabasz, J

doi:10.1080/03610927408827101

Cited by 4,947 publications

(3,099 citation statements)

References 7 publications

Supporting

Mentioning

3,033

Contrasting

Unclassified

Order By: Relevance

“…As noted above (Section 6.3.1), we used the Calinski & Harabasz (1974) variance ratio as a figure-of-merit. We used the NbClust package (Charrad et al 2012) for all the calculations, but this did not provide uncertainties…”

Section: Appendix a Kolmogorov-smirnov Statisticsmentioning

confidence: 99%

“…This involves choosing the partition that gives the highest ratio of the variance of the distances between objects in different clusters to the variance of distances between objects within clusters. Calinski & Harabasz (1974) plot the variance ratio as a function of the number of clusters and use the first local maximum to define the best number of clusters. We modified this criterion by requiring that the selected peak was significantly (3 standard deviations) above the neighbouring points 1 .…”

Section: Application Of the Clustering Algorithmmentioning

confidence: 99%

“…For a given number of clusters, this finds the cluster positions that minimise the sum of the squares of the distances from each data point to its cluster centre. We determined the best number of clusters to choose by using the Calinski & Harabasz (1974) variance ratio criterion (as implemented in the NbClust package Charrad et al 2012). This involves choosing the partition that gives the highest ratio of the variance of the distances between objects in different clusters to the variance of distances between objects within clusters.…”

Section: Application Of the Clustering Algorithmmentioning

confidence: 99%

See 2 more Smart Citations

Galaxy And Mass Assembly (GAMA): the unimodal nature of the dwarf galaxy population

Mahajan

Drinkwater

Driver

et al. 2014

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

In this paper we aim to (i) test the number of statistically distinct classes required to classify the local galaxy population, and, (ii) identify the differences in the physical and star formation properties of visually-distinct galaxies. To accomplish this, we analyse the structural parameters (effective radius (R ef f ), effective surface brightness within R ef f ( µ e ), central surface brightness (µ 0 ), and Sérsic index (n)), obtained by fitting the light profile of 432 galaxies (0.002 < z 0.02; Viking Z-band), and their spectral energy distribution using multi-band photometry in 18 broadbands to obtain the stellar mass (M * ), the star formation rate (SFR), the specific SFR (sSFR) and the dust mass (M dust ), respectively.We show that visually distinct, star-forming dwarf galaxies (irregulars, blue spheroids and low surface brightness galaxies) form a unimodal population in a parameter space mapped by µ e , µ 0 , n, R ef f , SFR, sSFR, M * , M dust and (g − i). The SFR and sSFR distribution of passively evolving (dwarf) ellipticals on the other hand, statistically distinguish them from other galaxies with similar luminosity, while the giant galaxies clearly segregate into starforming spirals and passive lenticulars. We therefore suggest that the morphology classification scheme(s) used in literature for dwarf galaxies only reflect the observational differences based on luminosity and surface brightness among the apparent distinct classes, rather than any physical differences between them.

show abstract

Section: Appendix a Kolmogorov-smirnov Statisticsmentioning

confidence: 99%

Section: Application Of the Clustering Algorithmmentioning

confidence: 99%

Section: Application Of the Clustering Algorithmmentioning

confidence: 99%

See 1 more Smart Citation

Galaxy And Mass Assembly (GAMA): the unimodal nature of the dwarf galaxy population

Mahajan

Drinkwater

Driver

et al. 2014

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

show abstract

“…The original and adaptive AP methods are implemented using the MATLAB program obtained from [19]. The performances of the three algorithms are evaluated using the Calinski-Harabasz criterion [20], which is the ratio of the between-cluster variance to the total within-cluster variance, defined as…”

Section: Improved Affinity Propagation (Iap) Performancementioning

confidence: 99%

Sound-Environment Monitoring Method Based on Computational Auditory Scene Analysis

Kawamoto¹

2017

JSIP

View full text Add to dashboard Cite

Monitoring techniques are a key technology for examining the conditions in various scenarios, e.g., structural conditions, weather conditions, and disasters. In order to understand such scenarios, the appropriate extraction of their features from observation data is important. This paper proposes a monitoring method that allows sound environments to be expressed as a sound pattern. To this end, the concept of synesthesia is exploited. That is, the keys, tones, and pitches of the monitored sound are expressed using the three elements of color, that is, the hue, saturation, and brightness, respectively. In this paper, it is assumed that the hue, saturation, and brightness can be detected from the chromagram, sonogram, and sound spectrogram, respectively, based on a previous synesthesia experiment. Then, the sound pattern can be drawn using color, yielding a "painted sound map." The usefulness of the proposed monitoring technique is verified using environmental sound data observed at a galleria.

show abstract

“…The index of Calinski & Harabasz (CH) and the internal measure of cohesion of the sum of the 553 squares within the group (WSS) are selected to this end [96][97][98]. The stop rule is the value closest to 554 the area where the curves interact.…”

mentioning

confidence: 99%