Paper Title
Dimensionality Reduction for Image Analysis

Abstract
In this article, we consider dimensionality reduction methods for image processing. Various dimensionality reduction methods exist in the literature, but we only consider the unsupervised statistical method of Principal Component Analysis (PCA). We show how the dimensionality of an image can be reduced via PCA without losing resolution of the image. For this purpose, a jpeg image of a hummingbird is downloaded and read in R. PCA is next performed of the large image dataset and the mean square error (mse) method is used for selection of number of components retained. This article shows that a clear image of the hummingbird can be obtained by just using first three or four PC-scores. Keywords - Dimensionality Reduction, Intensity, Principle Component Analysis, Variation