Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorBhardwaj, Pavika-
dc.contributor.authorKumar, Pardeep [Guided by]-
dc.contributor.authorKumar, Yugal [Guided by]-
dc.description.abstractMachine learning is a set of techniques which allow a machine to act as human beings. There are various essential components of machine learning knowledge pyramid. It includes symbols, facts, data, information, knowledge, intelligence and wisdom. Machine learning algorithms are searched with optimization. They are predictive in nature. They are least dependent on the user. They are applied on huge collection of data. Data mining act as an application of it. Though K-Means is the simplest technique of clustering to be used, still it has certain drawbacks. This project mainly deals with using harris-hawk meta-heuristic optimization technique in clustering. They provide an edge over traditional partitioning techniques because of its successful implementation and high intensity. The project aims to obtain optimized cluster centres. “Hawks” represent the number of clusters needed. “Location of the rabbits” are represented as initial and final cluster centres. The proposed algorithm is further evaluated on two parameters namely accuracy and intra-cluster distance. It leads to high accuracy and low intra-cluster distance.en_US
dc.publisherJaypee University of Information Technology, Solan, H.P.en_US
dc.subjectMachine learningen_US
dc.subjectOptimization algorithmsen_US
dc.subjectSoftware requirementsen_US
dc.titleDesign a New Clustering Algorithm for Partition Clustering Problemen_US
dc.typeProject Reporten_US
Appears in Collections:Dissertations (M.Tech.)

Files in This Item:
File Description SizeFormat 
Design a New Clustering Algorithm for Partition Clustering Problem.pdf2.36 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.