We propose a framework and a complete implementation of a translation and scale invariant image recognition system for natural indoor scenes. The system employs higher order autocorrelation features of scale space data which permit linear classification. An optimal linear classification method is presented, which is able to cope with a large number of classes represented by many, as well as very few samples. In the course of the analysis of our system, we examine which numerical methods for feature transformation and classification show sufficient stability to fulfill these demands. The implementation has been extensively tested. We present the results of our own application and several classification benchmarks. Image recognition Face recognition Scale invariancy Scale space Higher order autocorrelation Optimal linear classification 1. INTRODUCTION The task of visual recognition which was defined by Marr (1) with the question: "What objects are where in the environment?" is still ...

In this paper we present a method to extract qualitative information from any classification model that uses decision regions to generalize (e.g., feedforward neural nets, SVMs, etc). The method's complexity is independent of the dimensionality of the input data or model, making it computationally feasible for the analysis of even very high-dimensional models. The qualitative information extracted by the method can be directly used to analyze the classification strategies employed by a model, and also to compare strategies across diferent model types.

Abstract. The performance of principal component analysis (PCA) suffers badly in the presence of outliers. This paper proposes two novel approaches for robust PCA based on semidefinite programming. The first method, maximum mean absolute deviation rounding (MDR), seeks directions of large spread in the data while damping the effect of outliers. The second method produces a low-leverage decomposition (LLD) of the data that attempts to form a low-rank model for the data by separating out corrupted observations. This paper also presents efficient computational methods for solving these SDPs. Numerical experiments confirm the value of these new techniques. 1.

. The RCE algorithm, which can be considered as belonging to the category of the Region of Influence (ROI) incremental algorithms, is one of the first and most widely used incremental algorithms. It is well known that the performance of this algorithm is limited by constraints affecting the initial parameter settings: order in which input patterns are presented during the training phase and initial radius to be used for the generated hyperspheres. In this paper, we propose a deterministic method for obtaining automatically the suitable initial radius for a given input problem. The method is based on the definition of several functions which give a quantitative measure of the geometrical relations between the different categories defined in the input space. As the simulation results show, this method improves significantly the performance (in terms of number of units generated and generalization capability) of the RCE algorithm. 1. Introduction In the last years there has been an increa...

Breast cancer, which is the most common cancer in women, is a complex disease characterised by multiple molecular alterations. Current routine clinical management relies on availability of robust clinical and pathologic prognostic and predictive factors, like the Nottingham Prognostic Index, to support decision making. Recent advances in highthroughput molecular technologies supported the evidence of a biologic heterogeneity of breast cancer. This thesis is a multi-disciplinary work involving both computer scientists and molecular pathologists. It focuses on the development of advanced computational models for the classification of breast cancer into sub-types of the disease based on protein expression levels of selected markers. In a previous study conducted at the University of Nottingham, it has been suggested that immunohistochemical analysis may be used to identify distinct biological classes of breast cancer. The objectives of this work were related both to the clinical and technical aspects. From a clinical point of view, the aim was to encourage a multiple techniques approach