Fast and efficient non-parametric classification and clustering methods for largr data sets

Show simple item record

dc.contributor.author Babu, V Suresh
dc.date.accessioned 2015-09-16T08:49:37Z
dc.date.available 2015-09-16T08:49:37Z
dc.date.issued 2009
dc.identifier.other ROLL NO.04610105
dc.identifier.uri http://gyan.iitg.ernet.in/handle/123456789/189
dc.description Supervisor: P Viswanath en_US
dc.description.abstract Pattern Classification and clustering are two prominent pattern recognition tasks applied in various domains. Non-parametric methods are those which does not assume any model or distribution from for the data. Hence these methods are more general and can give better results provided the data set is a larger one. Nearest neighbor classifier(NNC) and its variants like k nearest neighbor classifier (k-NNC) are popular non-parametric classifiers. They show good performance and has asymptotic behavior comparable to that of the bayes classifier. When it comes to clustering methods, DBSCAN(Density based spatial clustering of applications with noise) uses density which is found non-parametrically at a point in order to derive density based clusters. DBSCAN can find arbitrary shaped clusters(unlike methods like k-means clustering) along with noisy outliers detection... en_US
dc.language.iso en en_US
dc.relation.ispartofseries TH-0774;
dc.subject COMPUTER SCIENCE AND ENGINEERING en_US
dc.title Fast and efficient non-parametric classification and clustering methods for largr data sets en_US
dc.type Thesis en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search


Browse

My Account