This paper describes an experimental com- parison of three unsupervised learning algorithms that distinguish the sense of an ambiguous word in untagged text.

A critical issue of applying Linear Discriminant Analysis (LDA) is both the singularity and instability of the within-class scatter matrix. In practice, particularly in image recognition applications such as face recognition, there are often a large number of pixels or pre-processed features available, but the total number of training patterns is limited and commonly less than the dimension of the feature space. In this paper, a new LDA-based method is proposed. It is based on a straighforward stabilisation approach for the within-class scatter matrix. In order to evaluate its effectiveness, experiments on face recognition using the well-known ORL and FERET face databases were carried out and compared with other LDA-based methods. The results indicate that our method improves the LDA classification performance when the within-class scatter matrix is not only singular but also poorly estimated, with or without a Principal Component Analysis intermediate step and using less linear discriminant features.

Abstract not found

Principal Component Analysis (PCA) is an example of a successful unsupervised statistical dimensionality reduction method, especially in small sample size problems. Despite the well-known attractive properties of PCA, the traditional approach does not incorporate prior information extracted from a specific domain knowledge. The development of techniques that bring together dimensionality reduction and prior knowledge can be performed in the framework of supervised learning methods, like Fisher Discriminant Analysis. Semi-supervised methods can also be applied if only a small number of labeled samples is available. In this paper, we propose a simple and efficient supervised method that allows PCA to incorporate explicitly domain knowledge and generates an embedding space that inherits its optimality properties for dimensionality reduction. The method relies on discriminant weights given by separating hyperplanes to generate the spatially weighted PCA. Several experiments using 2D frontal face images and different data sets have been carried out to illustrate the usefulness of the method for dimensionality reduction, classification and interpretation of face images. 1

Abstract not found