Clustering methods need to be robust if they are to be useful in practice. In this paper, we analyze several popular robust clustering methods and show that they have much in common. We also establish a connection between fuzzy set theory and robust statistics and point out the similarities between robust clustering methods and statistical methods such as the weighted least-squares (LS) techniq...

Regularized estimators in the context of group variables have been applied successfully in model and feature selection in order to preserve interpretability. We formulate a Distributionally Robust Optimization (DRO) problem which recovers popular estimators, such as Group Square Root Lasso (GSRL). Our DRO formulation allows us to interpret GSRL as a game, in which we learn a regression paramete...

In the context of dense stereo matching of pixels, we study the combination of different correlation measures. Considering the previous work about correlation measures, we use some measures that are the most significant in five kinds of measures based on: cross-correlation, classic statistics, image derivatives, nonparametric statistics and robust statistics. More precisely, this study validate...

In this paper, we extend a traditional robust datahiding set-up with host state at the encoder to a case when a partial side information about host statistics is also available at the decoder. We demonstrate that the knowledge of host statistics at the decoder can relax the critical requirements of random binning-based methods concerning attack channel statistics at the encoder. We also analyze...

Neural stimulus selectivity is thought to be optimized for the representation of real-world stimuli. Neural coding properties, therefore, may adapt to different environments. Here, we address the question if tuning curves depend on the statistics of visual stimuli. This is done by studying the directional tuning of macaque area MT neurons exposed to dynamic motion stimuli of two different direc...

Spatial structure of genetic variation within populations, an important interacting influence on evolutionary and ecological processes, can be analyzed in detail by using spatial autocorrelation statistics. This paper characterizes the statistical properties of spatial autocorrelation statistics in this context and develops estimators of gene dispersal based on data on standing patterns of gene...

In this paper, we extend a traditional robust data-hiding set-up with host state at the encoder to a case when a partial side information about host statistics is also available at the decoder. We demonstrate that the knowledge of host statistics at the decoder can relax the critical requirements of random binning-based methods concerning attack channel statistics at the encoder. We also analyz...

This thesis deals with the application of neural network and fuzzy logic based techniques for designing algorithm to recognize handwritten Bengali character. Fuzzy measures are related to methods of robust statistics, namely linear combinations of order statistics, which are used for combining information that may contain outliers, a common situation in handwriting recognition. Here first Chara...