Ronald Cools

Learn More
Despite their known weaknesses, hidden Markov models (HMMs) have been the dominant technique for acoustic modeling in speech recognition for over two decades. Still, the advances in the HMM framework have not solved its key problems: it discards information about time dependencies and is prone to overgeneralization. In this paper, we attempt to overcome(More)
Interference effects on task performance in conflict situations might reflect real limitations in inhibitory capabilities or failures to fully or consistently utilize such capabilities in executive control of task performance. We propose that useful clues regarding the actual cause of interference effects may be obtained from examination of their robustness(More)
Various lines of evidence suggest that the striatum is implicated in cognitive flexibility. The neuropsychological evidence has, for the most part, been based on research with patients with Parkinson's disease, which is accompanied by chemical disruption of both the striatum and the prefrontal cortex. The present study examined this issue by testing(More)
We reformulate the original component-by-component algorithm for rank-1 lattices in a matrix-vector notation so as to highlight its structural properties. For function spaces similar to a weighted Korobov space, we derive a technique which has construction cost O(sn log(n)), in contrast with the original algorithm which has construction cost O(sn). Herein s(More)
The component-by-component construction algorithm constructs the generating vector for a rank-1 lattice one component at a time by minimizing the worst-case error in each step. This algorithm can be formulated elegantly as a repeated matrixvector product, where the matrix-vector product expresses the calculation of the worst-case error in that step. As was(More)
One of the best known low-discrepancy sequences, used by many practitioners, is the Halton sequence. Unfortunately, there seems to exist quite some correlation between the points from the higher dimensions. A possible solution to this problem is the so-called scrambling. In this paper, we give an overview of known scrambling methods, and we propose a new(More)
Lattice rules are a family of equal-weight cubature formulas for approximating highdimensional integrals. By now it is well established that good generating vectors for lattice rules having n points can be constructed component-by-component for integrands belonging to certain weighted function spaces, and that they can achieve the optimal rate of(More)
CUBPACK aims to offer a collection of re-usable code for automatic <i>n</i>-dimensional (<i>n</i> &#8805; 1) numerical integration of functions over a collection of regions, i.e., quadrature and cubature. The current version allows this region to consist of a union of <i>n</i>-simplices and <i>n</i>-parellellepids. The framework of CUBPACK is described as(More)
Constraint Programming (CP) is a high-level declarative programming paradigm in which problems are modeled by means of constraints on the problem variables that need to hold in all solutions to the problem. Many problems of high practical relevance can easily be described in terms of constraints. Example application areas include production planning and(More)