When Kurt Goedel layed the foundations of theoretical computer science in 1931, he also introduced essential concepts of the theory of Artificial Intelligence (AI). Although much of subsequent AI research has focused on heuristics, which still play a major role in many practical AI applications, in the new millennium AI theory has finally become a full-fledged formal science, with important optimality results for embodied agents living in unknown environments, obtained through a combination of theory a la Goedel and probability theory. Here we look back at important milestones of AI history, mention essential recent results, and speculate about what we may expect from the next 25 years, emphasizing the significance of the ongoing dramatic hardware speedups, and discussing Goedel-inspired, self-referential, self-improving universal problem solvers.

This article presents a technique for proving problems hard for classes of the polynomial hierarchy or for PSPACE. The rationale of this technique is that some problem restrictions are able to simulate existential or universal quantifiers. If this is the case, reductions from Quantified Boolean Formulae (QBF) to these restrictions can be transformed into reductions from QBFs having one more quantifier in the front. This means that a proof of hardness of a problem at level n in the polynomial hierarchy can be split into n separate proofs, which may be simpler than a proof directly showing a reduction from a class of QBFs to the considered problem.

This paper proposes a neuro-rough model based on multi-layered perceptron and rough set. The neuro-rough model is then tested on modelling the risk of HIV from demographic data. The model is formulated using Bayesian framework and trained using Monte Carlo method and Metropolis criterion. When the model was tested to estimate the risk of HIV infection given the demographic data it was found to give the accuracy of 62%. The proposed model is able to combine the accuracy of the Bayesian MLP model and the transparency of Bayesian rough set model.

We try a conceptual analysis of inheritance diagrams, first in abstract terms, and then compare to "normality" and the "small/big sets" of preferential and related reasoning. The main ideas are about nodes as truth values and information sources, truth comparison by paths, accessibility or relevance of information by paths, relative normality, and prototypical reasoning.

In this paper, we address the inverse problem of reconstructing a scene as well as the camera motion from the image sequence taken by an omni-directional camera. Our structure from motion results give sharp conditions under which the reconstruction is unique. For example, if there are three points in general position and three omni-directional cameras in general position, a unique reconstruction is possible up to a similarity. We then look at the reconstruction problem with m cameras and n points, where n and m can be large and the over-determined system is solved by least square methods. The reconstruction is robust and generalizes to the case of a dynamic environment where landmarks can move during the movie capture. Possible applications of the result are computer assisted scene reconstruction, 3D scanning, autonomous robot navigation, medical tomography and city reconstructions.

We present and analyse three online algorithms for learning in discrete Hidden Markov Models (HMMs) and compare them with the Baldi-Chauvin Algorithm. Using the Kullback-Leibler divergence as a measure of generalisation error we draw learning curves in simplified situations. The performance for learning drifting concepts of one of the presented algorithms is analysed and compared with the Baldi-Chauvin algorithm in the same situations. A brief discussion about learning and symmetry breaking based on our results is also presented.

We consider the generic regularized optimization problem $\hat{\mathsf{\beta}}(\lambda)=\arg \min_{\beta}L({\sf{y}},X{\sf{\beta}})+\lambda J({\sf{\beta}})$. Efron, Hastie, Johnstone and Tibshirani [Ann. Statist. 32 (2004) 407--499] have shown that for the LASSO--that is, if $L$ is squared error loss and $J(\beta)=\|\beta\|_1$ is the $\ell_1$ norm of $\beta$--the optimal coefficient path is piecewise linear, that is, $\partial \hat{\beta}(\lambda)/\partial \lambda$ is piecewise constant. We derive a general characterization of the properties of (loss $L$, penalty $J$) pairs which give piecewise linear coefficient paths. Such pairs allow for efficient generation of the full regularized coefficient paths. We investigate the nature of efficient path following algorithms which arise. We use our results to suggest robust versions of the LASSO for regression and classification, and to develop new, efficient algorithms for existing problems in the literature, including Mammen and van de Geer's locally adaptive regression splines.

We propose a special computational device which uses light rays for solving the subset-sum problem. The device has a graph-like representation and the light is traversing it by following the routes given by the connections between nodes. The nodes are connected by arcs in a special way which lets us to generate all possible subsets of the given set. To each arc we assign either a number from the given set or a predefined constant. When the light is passing through an arc it is delayed by the amount of time indicated by the number placed in that arc. At the destination node we will check if there is a ray whose total delay is equal to the target value of the subset sum problem (plus some constants).

For binary classification we establish learning rates up to the order of $n^{-1}$ for support vector machines (SVMs) with hinge loss and Gaussian RBF kernels. These rates are in terms of two assumptions on the considered distributions: Tsybakov's noise assumption to establish a small estimation error, and a new geometric noise condition which is used to bound the approximation error. Unlike previously proposed concepts for bounding the approximation error, the geometric noise assumption does not employ any smoothness assumption.

This is to present work on modifying the Aleph ILP system so that it evaluates the hypothesised clauses in parallel by distributing the data-set among the nodes of a parallel or distributed machine. The paper briefly discusses MPI, the interface used to access message- passing libraries for parallel computers and clusters. It then proceeds to describe an extension of YAP Prolog with an MPI interface and an implementation of data-parallel clause evaluation for Aleph through this interface. The paper concludes by testing the data-parallel Aleph on artificially constructed data-sets.