Unsupervised Learning of Manifolds via Linear Approximations

In this paper, we examine the application of manifold learning to the clustering problem. The method used is Locality Preserving Projections (LPP), which is chosen because of its computational ef ciency. A detailed derivation of the method is presented, as well as the theoretical justication behind it. Experiments performed on CMU's PIE database show that the projections created by LPP yield better clustering results than those obtained by k-means alone.

• Type: Conference
• Year: 2007
• Where: DEXAW
• Authors: Hassan A. Kingravi, M. Emre Celebi, Pragya P. Rajauria
• Keywords: Clustering Problem, Computational Ef Ciency, Database, DEXAW 2007, Locality Preserving Projections