MNV for Clustering based on Non Symmetric Symbolic Proximity
Bapu B. Kiranagi, D. S. Guru, S. Manjunath, Bukahally Somashekar Harish · MyPrints@UOM (Mysore University Library) · 2009
Abstract — In this paper, we bring out the importance of nonsymmetric proximity values among symbolic objects in simulating the reality during clustering. The concept of Mutual Neighborhood Value (MNV) has been exploited on non-symmetric proximity values. The results of the experiments conducted reveal that the approaches based on non-symmetric proximity measures best suite the reality than the symmetric proximity measures. Index Terms — symbolic objects, non-symmetric proximity values, mutual neighborhood value, clustering. I.