Similaarity C-Means Clustering Algorithm
碩士 === 中原大學 === 數學系 === 88 === We develop a simple and effective approach to clustering which is called the similarity c-means clustering algorithm. This algorithm is an objective function based clustering method by maximizing the total similarity. The memberships resulting from the will-known fuzzy...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2000
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/69664519146095922902 |
| Summary: | 碩士 === 中原大學 === 數學系 === 88 === We develop a simple and effective approach to clustering which is called the similarity c-means clustering algorithm. This algorithm is an objective function based clustering method by maximizing the total similarity. The memberships resulting from the will-known fuzzy c-means clustering and its derivatives, however, do not always correspond to the explanation of degree of belonging of the data and has trouble under noisy environment
. In this paper, we will show that the similarity c-means algorithm have high ability of detecting noise and also have more reasonable and more possibilistic memberships.
|
|---|