A new suboptimal search strategy for feature selection is presented. It represents a more sophisticated version of "classical" floating search algorithms (Pudil et al., 1994), attempts to remove some of their potential deficiencies and facilitates finding a solution even closer to the optimal one.

this paper is to investigate the applicability of these techniques to high dimensional problems of Feature Selection. The aim is to establish whether the properties inferred for these techniques from medium scale experiments involving up to a few tens of dimensions extend to dimensionalities of one order of magnitude higher. Further, relative merits of these techniques vis-a-vis such high dimensional problems are explored and the possibility of exploiting the best aspects of these methods to create a composite feature selection procedure with superior properties is considered

The vast majority of classification systems are designed with a single set of features, and optimised to a single specified cost. However, in examples such as medical and financial risk modelling, costs are known to vary subsequent to system design. In this paper, we present a design method for feature selection in the presence of varying costs. Starting from the Wilcoxon nonparametric statistic for the performance of a classification system, we introduce a concept called the maximum realisable receiver operating characteristic (MRROC), and prove a related theorem. A novel criterion for feature selection, based on the area under the MRROC curve, is then introduced. This leads to a framework which we call Parcel. This has the flexibility to use different combinations of features at different operating points on the resulting MRROC curve. Empirical support for each stage in our approach is provided by experiments on real world problems, with Parcel achieving superior results. iv v C...

1.3 El problema de optimización..................... 6

Abstract. Texture representation of ultrasound (US) images is currently considered a major issue in medical image analysis. This paper investigates the texture representation of thyroid tissue via features based on the Contourlet Transform (CT) using different types of filter banks. A variety of statistical texture features based on CT coefficients, have been considered through a selection schema. The Sequential Float Feature Selection (SFFS) algorithm with a k-NN classifier has been applied in order to investigate the most representative set of CT features. For the experimental evaluation a set of normal and nodular ultrasound thyroid textures have been utilized. The maximum classification accuracy was 93%, showing that CT based texture features can be successfully applied for the representation of different types of texture in US thyroid images.