In the context of clustering, we consider a generative model in a Euclidean ambient space with clusters of different shapes, dimensions, sizes and densities. In an asymptotic setting where the number of points becomes large, we obtain theoretical guaranties for a few emblematic methods based on pairwise distances: a simple algorithm based on the extraction of connected components in a neighborhood graph; the spectral clustering method of Ng, Jordan and Weiss; and hierarchical clustering with single linkage. The methods are shown to enjoy some near-optimal properties in terms of separation between clusters and robustness to outliers. The local scaling method of Zelnik-Manor and Perona is shown to lead to a near-optimal choice for the scale in the first two methods. We also provide a lower bound on the spectral gap to consistently choose the correct number of clusters in the spectral method.

We investigate the issue of model selection and the use of the nonconformity (strangeness) measure in batch learning. Using the nonconformity measure we propose a new training algorithm that helps avoid the need for Cross-Validation or Leave-One-Out model selection strategies. We provide a new generalisation error bound using the notion of nonconformity to upper bound the loss of each test example and show that our proposed approach is comparable to standard model selection methods, but with theoretical guarantees of success and faster convergence. We demonstrate our novel model selection technique using the Support Vector Machine.

We propose Interactive Differential Evolution (IDE) based on paired comparisons for reducing user fatigue and evaluate its convergence speed in comparison with Interactive Genetic Algorithms (IGA) and tournament IGA. User interface and convergence performance are two big keys for reducing Interactive Evolutionary Computation (IEC) user fatigue. Unlike IGA and conventional IDE, users of the proposed IDE and tournament IGA do not need to compare whole individuals each other but compare pairs of individuals, which largely decreases user fatigue. In this paper, we design a pseudo-IEC user and evaluate another factor, IEC convergence performance, using IEC simulators and show that our proposed IDE converges significantly faster than IGA and tournament IGA, i.e. our proposed one is superior to others from both user interface and convergence performance points of view.

Principal component analysis (PCA) is an essential tool for data analysis and unsupervised dimensionality reduction, whose goal is to find, among linear combinations of the data variables, a sequence of orthogonal factors that most efficiently explain the variance of the observations. One of its main shortcomings is that, even if PCA finds a small number of important factors, the factor themselves typically involve all original variables. In the last decade, several alternatives to PCA which find sparse and potentially interpretable factors have been proposed, notably nonnegative matrix factorization (NMF) [2] and sparse PCA (SPCA) [3, 4, 5]. However, in many applications, only constraining the size of the factors does not seem appropriate because the considered factors are not only expected to be sparse but also to have a certain structure. In fact, the popularity of NMF for face image analysis owes essentially to the fact that the method happens to retrieve sets of variables that are localized on the face and capture some features or parts of the face which seem intuitively meaningful given our a priori.

The problem of ranking a set of objects given some measure of similarity is one of the most basic in machine learning. Recently Agarwal proposed a method based on techniques in semi-supervised learning utilizing the graph Laplacian. In this work we consider a novel application of this technique to ranking binary choice data and apply it specifically to ranking US Senators by their ideology.

Grid environment is a service oriented infrastructure in which many heterogeneous resources participate to provide the high performance computation. One of the bug issues in the grid environment is the vagueness and uncertainty between advertised resources and requested resources. Furthermore, in an environment such as grid dynamicity is considered as a crucial issue which must be dealt with. Classical rough set have been used to deal with the uncertainty and vagueness. But it can just be used on the static systems and can not support dynamicity in a system. In this work we propose a solution, called Dynamic Rough Set Resource Discovery (DRSRD), for dealing with cases of vagueness and uncertainty problems based on Dynamic rough set theory which considers dynamic features in this environment. In this way, requested resource properties have a weight as priority according to which resource matchmaking and ranking process is done. We also report the result of the solution obtained from the simulation in GridSim simulator. The comparison has been made between DRSRD, classical rough set theory based algorithm, and UDDI and OWL S combined algorithm. DRSRD shows much better precision for the cases with vagueness and uncertainty in a dynamic system such as the grid rather than the classical rough set theory based algorithm, and UDDI and OWL S combined algorithm.

Regularized risk minimization with the binary hinge loss and its variants lies at the heart of many machine learning problems. Bundle methods for regularized risk minimization (BMRM) and the closely related SVMStruct are considered the best general purpose solvers to tackle this problem. It was recently shown that BMRM requires $O(1/\epsilon)$ iterations to converge to an $\epsilon$ accurate solution. In the first part of the paper we use the Hadamard matrix to construct a regularized risk minimization problem and show that these rates cannot be improved. We then show how one can exploit the structure of the objective function to devise an algorithm for the binary hinge loss which converges to an $\epsilon$ accurate solution in $O(1/\sqrt{\epsilon})$ iterations.

This paper introduces a Bayesian framework to detect multiple signals embedded in noisy observations from a sensor array. For various states of knowledge on the communication channel and the noise at the receiving sensors, a marginalization procedure based on recent tools of finite random matrix theory, in conjunction with the maximum entropy principle, is used to compute the hypothesis selection criterion. Quite remarkably, explicit expressions for the Bayesian detector are derived which enable to decide on the presence of signal sources in a noisy wireless environment. The proposed Bayesian detector is shown to outperform the classical power detector when the noise power is known and provides very good performance for limited knowledge on the noise power. Simulations corroborate the theoretical results and quantify the gain achieved using the proposed Bayesian framework.

Existing solutions are very limited (chats, forums...). Many sites exist, for Web-users to express their opinion, but can they build structured arguments and confront them? We try to answer these questions by offering new software based on logic. But then, what kind of logic can describe the arguments of a debate? What is the role of n-opposition theory? In this paper we compare the results of three experiments. First we will introduce some important notions, then we will present our experimental protocol and finally we will present detailed discussion system. We will also explain how we use the n-opposition theory to structure debate and reason. To conclude, we will discuss experiments contexts and results.

We describe in this article a multiagent urban traffic simulation, as we believe individual-based modeling is necessary to encompass the complex influence the actions of an individual vehicle can have on the overall flow of vehicles. We first describe how we build a graph description of the network from purely geometric data, ESRI shapefiles. We then explain how we include traffic related data to this graph. We go on after that with the model of the vehicle agents: origin and destination, driving behavior, multiple lanes, crossroads, and interactions with the other vehicles in day-to-day, ?ordinary? traffic. We conclude with the presentation of the resulting simulation of this model on the Rouen agglomeration.