Some Extensions of Buser and Baroni-Urbani's Clustering Method

✍ Scribed by L. Dragomirescu

• Publisher: John Wiley and Sons
• Year: 1987
• Tongue: English
• Weight: 382 KB
• Volume: 29
• Category: Article
• ISSN: 0323-3847
• DOI: 10.1002/bimj.4710290317

✦ Synopsis

We consider that thealgorithm of numerical taxonomy focused upon the concept of "homogeneity", proposed by BIJSEB and BABONI-URBANI (1982) for binary data, satisfies beat from all the other algorithms known by us, the conditions of claeeification in biology formulated by BECKNEB in 1959.

Therefore the paper proposes and studies hom eneitiea for binary and for ordered multi-states data, which model explicitly Becher's conditio8. From all these, the homogeneity denoted by h* has the remarkable quality of generalizing the famous Jaccard's similarity coefficient. Moreover, using h* in thia algorithm ''naturally'' improved by us, we can achieve the performance of a correct classification of the remarkable example proposed by WATANABE (1969).