Intelligent Data Analysis: An International Journal

We know that there are many clustering methods for the case of a known/unknown number of clusters. Clustering is a result of fulfillment of some stopping criterion. Usually, optimisation of some quality criterion is performed or iterative processes are accomplished. How to estimate the quality of clustering obtained by some method? Is the obtained clustering result corresponding to the objective reality or some stopping criterion of the algorithm is made and we have obtained only some partition? Here, a practical approach and the common general criteria based on an estimation of the stability of clustering are submitted. The criterion does not use any probabilistic assumptions or distances in feature space. For some well-known clustering algorithms, efficient methods for computing the introduced stability criteria according to the training set are obtained. Some illustrative real and artificial examples for various situations are shown.

The forensic detection of median filtering has recently attracted the attention of the research community, mainly because of the median filtering potential uses for tampering and concealing image tampering traces in digital images. In this paper, we propose multi-scale and multi-perturbation solutions that build a highly discriminative feature space, which highlights the artifacts of median filtering by means of image quality measures. The proposed methods achieve promising results when validated with a series of real-world test cases, comprising different image compression levels, resolutions, and also a cross-dataset validation protocol.

k-NN is one of the most popular and effective classifiers nowadays. However, it has some limitations that overcome its applicability in large scale scenarios: basically, it requires storing the whole training set, and it computes distances of a test sample with the training data set. These limitations have been traditionally alleviated with data reduction techniques. This paper introduces a multi-objective evolutionary approach for data reduction. Our method simultaneously generates prototypes and selects features for k-NN classifiers. Contrary to most of the existing approaches, our method treats the problem with multi-objective evolutionary optimizers. We show the effectiveness of our proposal in benchmark data and compare its performance with state of the art techniques.

DNA microarrays is a technology that can be used to diagnose cancer and other diseases. To automate the analysis of such data, pattern recognition and machine learning algorithms can be applied. However, the curse of dimensionality is unavoidable: very few samples to train, and many attributes in each sample. As the predictive accuracy of supervised classifiers decays with irrelevant and redundant features, the necessity of a dimensionality reduction process is essential. The main idea is to retain only the genes that are the most influential in the classification of the disease. In this paper, a new methodology based on Principal Component Analysis and Logistics Regression is proposed. Our method enables the selection of particular genes that are relevant for classification. Experiments were run using eight different classifiers on two benchmark datasets: Leukemia and Lymphoma. The results show that our method not only reduces the number of required attributes, but also increase the classification accuracy in more than 10% in all the cases we tested.

In this work the authors introduce a novel geometric voting scheme that extends previous algorithms, like Hough transform and tensor voting, in order to tackle perceptual organization problems. Our approach is grounded in three methodologies: representation of information using Conformal Geometric Algebra, a local voting process, which introduce global perceptual considerations at low level, and a global voting process, which clusters salient geometric entities that are supported in the whole image. Geometric algebra provides a suitable mathematical framework, which allows our algorithm to infer high-level geometric representations of percepts in an image. Experiments show the capability of our algorithm for representing objects in images in terms of circles and lines, even though it contains a noisy input, incomplete data, illusory or non-linear contours.

In recent years a simple representation of a speech excerpt has been proposed, as a binary matrix allowing easy access to the speaker discriminant information. In addition to the time-related abilities of this representation, it also allows the system to work with a temporal information representation based on sequential changes present in the binary representation. A new temporal information is proposed in order to add it to speaker recognition systems. A new specificity selection approach using a mask in the cumulative vector space is also proposed. Furthermore in this space, temporal information can be exploited to compensate for the effects of session variability. A new variability compensation method in the temporal space is proposed in order to remove the unwanted attributes of session variability and the common attributes among speakers. This aims to increase effectiveness in the speaker binary key paradigm. The experimental validation, done on the NIST-SRE framework, demonstrates the efficiency of the proposed solutions, which shows an EER improvement of 9%. The combination of i-vector and binary approaches, using the proposed methods, showed the complementarity of the discriminatory information exploited by each of them.

Models based on q-grams are widely used in communication theory, natural language processing, statistical pattern classification, and other areas of machine learning. In this paper, the idea ofgeometric q-grams is used to solve the place recognition problem, which is fundamental for autonomous navigation. Visual place recognition implies the capacity to recognize a previously visited place given current camera measurements (i.e. digital images). A conventional approach for solving the place recognition problem is to employ histograms of visual features, typically losing the spatial location of said features (relative to the image) in the process. Our approach first detects interest points and assigns labels to them based on visual features; then the images are represented by a set of geometric q$-grams obtained from triangles of anexpanded Delaunay triangulation, thus implicitly encoding relative spatial information. Through this representation and the use of an inverted index, images that match a query can be efficiently recovered in real time. The proposed approach is path independent and was tested on publicly available datasets, resulting in a high recall rate and reduced time complexity.

This paper is concerned with the Quaternion Support Vector Machines for classification as a generalization of the real- and complex-valued Support Vector Machines. In this framework we handle the design of kernels involving the Clifford or quaternion product. The application section shows experiments in pattern recognition and colour image processing. We also present a way to expand the amount of classes without the need to increase the number of classifiers as in standar approaches.

It has been reported in many works on skin detection and segmentation from color images that skin color models suffer from low specificity and high variance of the skin color, and this problem can be addressed by conforming the skin model to a presented scene. Here, we introduce a new hybrid adaptation system which combines two strategies, namely (i) adaptation from a detected facial region and (ii) a self-adaptive scheme that creates a local model based on the response obtained using the global one. As a result of this hybrid adaptation, we obtain a local skin color model and we use it to extract seeds for the geodesic distance transform that determines the boundaries of skin regions. The results of our extensive experimental study confirm that the proposed algorithm outperforms several state-of-the-art methods, as well as our earlier adaptive skin detectors.

Tracking-by-detection methods have become increasingly popular recently. This work presents a new multi-face tracking algorithm based on the association of detection responses given by a spatio-temporal face detector; which are considered as initial small trajectories or tracklets. An appearance model based on the spatio-termporal information is used to guide the tracker. Besides, a new adaptive Kalman filter that dynamically adjusts its parameters on the basis of the quality of the detector output is proposed. The introduced approach is evaluated on several challenging video sequences from the YouTube Faces database, achieving a very good performance.