作者: L. E. RENDELL , H. WHITEHEAD
DOI: 10.1080/09524622.2003.9753513
关键词:
摘要: ABSTRACT A common task for researchers of animal vocalisations is statistically comparing repertoires, or sets vocalisations. We evaluated five methods repertoires ‘codas’, short repeated patterns clicks, recorded from sperm whale (Physeter macrocephalus) groups. Three the involved classification codas—human observer classification, k-means cluster analysis using Calinski and Harabasz's (1974) criterion to determine k, a divisive clustering procedure Duda Hart's (1973) k. Two other used multivariate distances calculate similarity measures between coda repertoires. When on sample dataset, failed produce consistent results. did not provide clear signal determining number classes (k). Divisive performed satisfactorily and, encouragingly, gave similar results similari...