Knowledge Discovery and Data Mining (KDD) is an interdisciplinary area focusing on methodologies for extracting useful knowledge from data. Patterns of relations of data and information have the capacity to signify knowledge. Image pattern collection and management is the hottest subject of the digital world. The demand for image recognition knowledge of various kinds of real world images becomes greater. Kohenen's Self Organizing Maps (SOM) algorithm is one of the particular neural network algorithms, which is used for pattern learning and retrieval. The conventional SOM learning method represents poor knowledge and hence their applicable targets are restricted. In this paper SOM is scrutinizing with various standard distances and remarkable similar measures. Reliable image learning is achieved with City block, Lee, Maximum value distance, Jaccard and Dice coefficient. Image gallery can be mined well by using SOM with the above said measures. The composed knowledge is useful for various significant services. © Springer-Verlag Berlin Heidelberg 2010.
CITATION STYLE
Chenthalir Indra, N., & Rama Raj, E. (2010). Similar - Dissimilar Victor Measure Analysis to Improve Image Knowledge Discovery Capacity of SOM. In Communications in Computer and Information Science (Vol. 101, pp. 389–393). https://doi.org/10.1007/978-3-642-15766-0_61
Mendeley helps you to discover research relevant for your work.