Paper Info
Title
Graph-Based Multilevel Dimensionality Reduction with Applications to Eigenfaces and Latent Semantic Indexing
Abstract
Dimension reduction techniques have been successfully applied to face recognition and text information retrieval. The process can be time-consuming when the data set is large. This paper presents a multilevel framework to reduce the size of the data set, prior to performing dimension reduction. The algorithm exploits nearest-neighbor graphs.It recursively coarsens the data by finding a maximal matching level by level.The coarsened data at the lowest level is then projected using a known linear dimensionality reduction method. The same linear mapping %as that of the lowest level is performed on the original data set, and on any new test data.The methods are illustrated on two applications: Eigenfaces (face recognition) and Latent Semantic Indexing (text mining). Experimental results indicate that the multilevel techniques proposed here %in this paper offer a very appealing cost to quality ratio.
Year
DOI
Venue
2008
10.1109/ICMLA.2008.140
ICMLA
Keywords
Field
DocType
original data,latent semantic indexing,dimension reduction,new test data,face recognition,lowest level,graph-based multilevel dimensionality reduction,dimension reduction technique,linear dimensionality reduction method,maximal matching level,coarsened data,text analysis,graph partitioning,principal component analysis,data reduction,face,information retrieval,nearest neighbor graph,eigenfaces,dimensionality reduction,linear mapping,text mining,databases,graph theory,indexing
Data mining,Eigenface,Dimensionality reduction,Computer science,Search engine indexing,Artificial intelligence,Graph theory,Facial recognition system,Pattern recognition,Matching (graph theory),Test data,Machine learning,Data reduction
Conference
Citations 
PageRank 
References 
9
0.64
11
Authors
3
Name
Order
Citations
PageRank
Sophia Sakellaridi1151.07
Haw-ren Fang213213.24
Yousef Saad31940254.74