Cluster optimization using appropriate nearest neighbour

ISSN: 22498958
0Citations
Citations of this article
2Readers
Mendeley users who have this article in their library.

Abstract

In this postulation, presents the clustering of facial images using machine learning algorithm such as nearest neighbor and approximate rank order clustering. Clustering is a technique for classifying similar kind of object based on their trait. Clustering of images is challenging problems and there is still a considerable measure of work that needs to be done in this area. Problems in clustering large dataset is to choose the quantity of clusters and evaluating the obtained clusters. Clustering regard as the most important unsupervised learning as it manages finding a structure in an accumulation of unlabeled information. A loose meaning of clustering could be "the way toward sorting out articles into clusters whose people are nearby one means or another. A cluster in this manner is an accumulation of items which are "comparable" amongst them and are "divergent" to the articles which are place with different cluster. This thesis presents a work to improve clustering method to decrease the number of clusters in a LFW (Labeled face in wild) dataset. Previous work uses kd tree a nearest neighbor method and approximate rank order clustering method to find cluster on LFW dataset. our proposed method implement ball tree a better nearest neighbor algorithm to reduce the number of clusters created by previous method.

References Powered by Scopus

Delving deep into rectifiers: Surpassing human-level performance on imagenet classification

15184Citations
8254Readers
Get full text

Robust Real-Time Face Detection

11087Citations
4080Readers
Get full text
7218Citations
5613Readers
Get full text

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Cite

CITATION STYLE

APA

Imam, N., & Tarar, S. (2018). Cluster optimization using appropriate nearest neighbour. International Journal of Engineering and Advanced Technology, 8(2), 114–121.

Readers' Seniority

Tooltip

Professor / Associate Prof. 1

50%

PhD / Post grad / Masters / Doc 1

50%

Readers' Discipline

Tooltip

Engineering 1

50%

Psychology 1

50%

Save time finding and organizing research with Mendeley

Sign up for free