The recent introduction of indefinite SVM by Luss and dAspremont [15] has effectively demonstrated SVM classification with a non-positive semi-definite kernel (indefinite kernel). This paper studies the properties of the objective function introduced there. In particular, we show that the objective function is continuously differentiable and its gradient can be explicitly computed. Indeed, we further show that its gradient is Lipschitz continuous. The main idea behind our analysis is that the objective function is smoothed by the penalty term, in its saddle (min-max) representation, measuring the distance between the indefinite kernel matrix and the proxy positive semi-definite one. Our elementary result greatly facilitates the application of gradient-based algorithms. Based on our analysis, we further develop Nesterovs smooth optimization approach [16,17] for indefinite SVM which has an optimal convergence rate for smooth problems. Experiments on various benchmark datasets validate our analysis and demonstrate the efficiency of our proposed algorithms.

This paper examines the generalization properties of online convex programming algorithms when the loss function is Lipschitz and strongly convex. Our main result is a sharp bound, that holds with high probability, on the excess risk of the output of an online algorithm in terms of the average regret. This allows one to use recent algorithms with logarithmic cumulative regret guarantees to achieve fast convergence rates for the excess risk with high probability. As a corollary, we characterize the convergence rate of PEGASOS (with high probability), a recently proposed method for solving the SVM optimization problem.

Networks are ubiquitous in science and have become a focal point for discussion in everyday life. Formal statistical models for the analysis of network data have emerged as a major topic of interest in diverse areas of study, and most of these involve a form of graphical representation. Probability models on graphs date back to 1959. Along with empirical studies in social psychology and sociology from the 1960s, these early works generated an active network community and a substantial literature in the 1970s. This effort moved into the statistical literature in the late 1970s and 1980s, and the past decade has seen a burgeoning network literature in statistical physics and computer science. The growth of the World Wide Web and the emergence of online networking communities such as Facebook, MySpace, and LinkedIn, and a host of more specialized professional network communities has intensified interest in the study of networks and network data. Our goal in this review is to provide the reader with an entry point to this burgeoning literature. We begin with an overview of the historical development of statistical network modeling and then we introduce a number of examples that have been studied in the network literature. Our subsequent discussion focuses on a number of prominent static and dynamic network models and their interconnections. We emphasize formal model descriptions, and pay special attention to the interpretation of parameters and their estimation. We end with a description of some open problems and challenges for machine learning and statistics.

This survey is an introduction to positive definite kernels and the set of methods they have inspired in the machine learning literature, namely kernel methods. We first discuss some properties of positive definite kernels as well as reproducing kernel Hibert spaces, the natural extension of the set of functions $\{k(x,\cdot),x\in\mathcal{X}\}$ associated with a kernel $k$ defined on a space $\mathcal{X}$. We discuss at length the construction of kernel functions that take advantage of well-known statistical models. We provide an overview of numerous data-analysis methods which take advantage of reproducing kernel Hilbert spaces and discuss the idea of combining several kernels to improve the performance on certain tasks. We also provide a short cookbook of different kernels which are particularly useful for certain data-types such as images, graphs or speech segments.

MontyLingua, an integral part of ConceptNet which is currently the largest commonsense knowledge base, is an English text processor developed using Python programming language in MIT Media Lab. The main feature of MontyLingua is the coverage for all aspects of English text processing from raw input text to semantic meanings and summary generation, yet each component in MontyLingua is loosely-coupled to each other at the architectural and code level, which enabled individual components to be used independently or substituted. However, there has been no review exploring the role of MontyLingua in recent research work utilizing it. This paper aims to review the use of and roles played by MontyLingua and its components in research work published in 19 articles between October 2004 and August 2006. We had observed a diversified use of MontyLingua in many different areas, both generic and domain-specific. Although the use of text summarizing component had not been observe, we are optimistic that it will have a crucial role in managing the current trend of information overload in future research.

We provide an overview of some recent progress on the complexity of election systems. The issues studied include the complexity of the winner, manipulation, bribery, and control problems.

The management and combination of uncertain, imprecise, fuzzy and even paradoxical or high conflicting sources of information has always been, and still remains today, of primal importance for the development of reliable modern information systems involving artificial reasoning. In this introduction, we present a survey of our recent theory of plausible and paradoxical reasoning, known as Dezert-Smarandache Theory (DSmT) in the literature, developed for dealing with imprecise, uncertain and paradoxical sources of information. We focus our presentation here rather on the foundations of DSmT, and on the two important new rules of combination, than on browsing specific applications of DSmT available in literature. Several simple examples are given throughout the presentation to show the efficiency and the generality of this new approach.

In this report, a novel approach to intelligence and learning is introduced; this approach is based upon what we called percep tion logic. W h at we call ' perception automata ' is introduced in which learning is accom p lished at different perception resolution. Learning in this autom a ta is not heuristic, rather it guarantees the convergence of the approxim a ted function to whatever precision required. Furthe rm ore, the learning process can take place on-line and in at m o st O(log(N)) epochs, where N is the num ber of sam p les. The perception autom a ta is based on hierarchal leve ls of resolution in which each level adds som e details to the constructed function till th e final level can successfully reconstruct the whole function. This approach com b ines the favors of com putational approach in the sense that it is precise, structural and rigorous, and the features of distributed processing and adaptivity of soft com puting, as well as continuity and real-tim e response of dynam i cal system s.

Instantaneously Trained Neural Networks Abhilash Ponnath Abstract: This paper presents a review of instantaneously trained neural networks (ITNNs). These networks trade learning time for size and, in the basic model, a new hidden node is created for each training sample. Various versions of the corner-classification family of ITNNs, which have f ound applications in artificial intelligence (AI), are described. Implementation issues are also considered. 1 Introduction The human brain, the most complex known living structure in the universe, has the nerve cell or neuron as its fundamental unit. The number of neurons and connections between the neurons is enormous; this ensemble enables the brain to surpass the computational capacity of supercomputers in existence today. Artificial neural networks (ANNs) are models of the brain, which implement the mapping, ƒ: X Y such that the task is completed in a "certain" sense.

The field of Artificial Intelligence has, for a long time, neglected the role of emotions in human cognition, with few but notable exceptions. This has been motivated in part by the assumption that the emulation of human rationality by a machine is sufficient for attaining general human-level intelligence. This paper reviews neuroscientific results showing empirical evidence, consistently for over a decade, sustaining that emotion mechanisms in the brain play a fundamental role in decision making processes, as well as in cognitive regulation. Moreover, this role takes place regardless of whether the subject is aware of any emotion. These mechanisms are particularly important in social contexts. Lesions in the pathways supporting these mechanisms provoke serious impairments on social behavior. For instance, subjects with lesions in the pathways between the orbitofrontal cortex and the amygdala are no longer able to sustain an healthy social live, despite their intact intellectual capabilities. Strikingly, these patients are even able to verbally describe what would be the proper social behavior, although are unable to follow it. One important mechanism in social contexts is empathy, fundamental for proper social relations. It has been proposed that empathy is founded on mechanisms analogous to the mirror neurons.