A specific type of neural network, the Restricted Boltzmann Machine (RBM), is implemented for classification and feature detection in machine learning. RBM is characterized by separate layers of visible and hidden units, which are able to learn efficiently a generative model of the observed data. We study a "hybrid" version of RBM's, in which hidden units are analog and visible units are binary, and we show that thermodynamics of visible units are equivalent to those of a Hopfield network, in which the N visible units are the neurons and the P hidden units are the learned patterns. We apply the method of stochastic stability to derive the thermodynamics of the model, by considering a formal extension of this technique to the case of multiple sets of stored patterns, which may act as a benchmark for the study of correlated sets. Our results imply that simulating the dynamics of a Hopfield network, requiring the update of N neurons and the storage of N(N-1)/2 synapses, can be accomplished by a hybrid Boltzmann Machine, requiring the update of N+P neurons but the storage of only NP synapses. In addition, the well known glass transition of the Hopfield network has a counterpart in the Boltzmann Machine: It corresponds to an optimum criterion for selecting the relative sizes of the hidden and visible layers, resolving the trade-off between flexibility and generality of the model. The low storage phase of the Hopfield model corresponds to few hidden units and hence a overly constrained RBM, while the spin-glass phase (too many hidden units) corresponds to unconstrained RBM prone to overfitting of the observed data.

An empirical investigation of the interaction of sample size and discretization - in this case the entropy-based method CAIM (Class-Attribute Interdependence Maximization) - was undertaken to evaluate the impact and potential bias introduced into data mining performance metrics due to variation in sample size as it impacts the discretization process. Of particular interest was the effect of discretizing within cross-validation folds averse to outside discretization folds. Previous publications have suggested that discretizing externally can bias performance results; however, a thorough review of the literature found no empirical evidence to support such an assertion. This investigation involved construction of over 117,000 models on seven distinct datasets from the UCI (University of California-Irvine) Machine Learning Library and multiple modeling methods across a variety of configurations of sample size and discretization, with each unique "setup" being independently replicated ten times. The analysis revealed a significant optimistic bias as sample sizes decreased and discretization was employed. The study also revealed that there may be a relationship between the interaction that produces such bias and the numbers and types of predictor attributes, extending the "curse of dimensionality" concept from feature selection into the discretization realm. Directions for further exploration are laid out, as well some general guidelines about the proper application of discretization in light of these results.

In this paper, we present some high level information fusion approaches for numeric and symbolic data. We study the interest of such method particularly for classifier fusion. A comparative study is made in a context of sea bed characterization from sonar images. The classi- fication of kind of sediment is a difficult problem because of the data complexity. We compare high level information fusion and give the obtained performance.

STA4273H : index

This work is motivated by the problem of image mis-registration in remote sensing and we are interested in determining the resulting loss in the accuracy of pattern classification. A statistical formulation is given where we propose to use data contamination to model and understand the phenomenon of image mis-registration. This model is widely applicable to many other types of errors as well, for example, measurement errors and gross errors etc. The impact of data contamination on classification is studied under a statistical learning theoretical framework. A closed-form asymptotic bound is established for the resulting loss in classification accuracy, which is less than $\epsilon/(1-\epsilon)$ for data contamination of an amount of $\epsilon$. Our bound is sharper than similar bounds in the domain adaptation literature and, unlike such bounds, it applies to classifiers with an infinite Vapnik-Chervonekis (VC) dimension. Extensive simulations have been conducted on both synthetic and real datasets under various types of data contamination, including label flipping, feature swapping and the replacement of feature values with data generated from a random source such as a Gaussian or Cauchy distribution. Our simulation results show that the bound we derive is fairly tight.

Functional data analysis involves data described by regular functions rather than by a finite number of real valued variables. While some robust data analysis methods can be applied directly to the very high dimensional vectors obtained from a fine grid sampling of functional data, all methods benefit from a prior simplification of the functions that reduces the redundancy induced by the regularity. In this paper we propose to use a clustering approach that targets variables rather than individual to design a piecewise constant representation of a set of functions. The contiguity constraint induced by the functional nature of the variables allows a polynomial complexity algorithm to give the optimal solution.

Most classification methods are based on the assumption that data conforms to a stationary distribution. The machine learning domain currently suffers from a lack of classification techniques that are able to detect the occurrence of a change in the underlying data distribution. Ignoring possible changes in the underlying concept, also known as concept drift, may degrade the performance of the classification model. Often these changes make the model inconsistent and regular updatings become necessary. Taking the temporal dimension into account during the analysis of Web usage data is a necessity, since the way a site is visited may indeed evolve due to modifications in the structure and content of the site, or even due to changes in the behavior of certain user groups. One solution to this problem, proposed in this article, is to update models using summaries obtained by means of an evolutionary approach based on an intelligent clustering approach. We carry out various clustering strategies that are applied on time sub-periods. To validate our approach we apply two external evaluation criteria which compare different partitions from the same data set. Our experiments show that the proposed approach is efficient to detect the occurrence of changes.

This page includes forthcoming AAAI sponsored conferences, conferences presented ICEIS 2012 will be held June 28 by AAAI Affiliates, and conferences held in cooperation with AAAI. RuleML-2012 will be AAAI Spring Symposium Series. The Thirteenth International Conference held August 27-31, 2012 in Montpellier, AAAI Spring Symposium Series will be on Principles of Knowledge France. KR The Third International Conference University, Stanford, California, USA 2012 will be held June 10-14, 2012 in on Computational Creativity. ICWSM-12 will be held June 4-7 at ICAPS 2012 will be held June 24-28, Twenty-Fifth International Conference Trinity College in Dublin, Ireland.

Modeling interaction with robots raises new and different challenges for dialog modeling than traditional dialog modeling with less embodied machines. We present four case studies of implementing a typical human-robot interaction scenario with different state-of-the-art dialog frameworks in order to identify challenges and pitfalls specific to HRI and potential solutions. The results are discussed with a special focus on the interplay between dialog and task modeling on robots.

While such systems have been shown to successfully support a broad range of interactions, they rely heavily on precoded data. For example, dialogue responses are typically limited to only one or two dozen phrases, which pales in comparison to the diversity of human speech. We believe that in order for robotic systems to become a truly ubiquitous technology, robots must make sense of natural human behavior and engage with humans in a more humanlike way. Robots must become more like humans instead of forcing humans to be more like robots. Much of human knowledge about the appropriateness of behavior, in terms of both speech and actions, comes from our personal experiences and our observations of others. We compare its performance variations form a knowledge base from which to a teleoperated robot following a scripted task we learn what to say and what actions to perform to protocol and examine both the behavior of the achieve certain goals.