We investigated the generalization capabilities of backpropagation learning in feed forward and re current feed forward connectionist networks on the assignment of syllable boundaries to ortho graphic representations in Dutch hyphenation This is a di cult task because phonological and morphological constraints interact leading to am biguity in the input patterns We compared the results to di er...

Model generation can be regarded as a spe­ cial case of the Constraint Satisfaction Problem (CSP). It has many applications in AI, com­ puter science and mathematics. In this paper, we describe SEM, a System for Enumerating fi­ nite Models of first-order many-sorted theories. To the best of our knowledge, SEM outperforms any other finite model generation system on many test problems. The high p...

This study builds on the earlier work by Odedina and Afullo on Multipath fading in Durban. Their work was based on multipath measurements in Durban over a 6.73 km Line-of-Sight (LOS) link. This submission uses the geoclimatic factor approach and ITU-R recommendations P530-14 to obtain the multipath fading occurrence in five cities in South Africa, including Durban. Three-year radiosonde data is...

The general problem of computing poste rior probabilities in Bayesian networks is NP hard Cooper However e cient algorithms are often possible for particular applications by exploiting problem struc tures It is well understood that the key to the materialization of such a possibil ity is to make use of conditional indepen dence and work with factorizations of joint probabilities rather than joi...

Notations /3, Prest a outer radius of the tube A apparatus constant t b distance to screen (in App. B) Tliq Bn 3 dB-bandwith B l o n g longitudinal magnetization of RL the tube gin, Vout C O ~ ]~Ni~ONi l velocity of propX agation of torsional waves in a XL tube without losses y C, C~, C2, C3 constants z deff thickness of "immobilized Z liquid layer" Z L f frequency a fn resonance frequency /~ F...

The paper explores the power of two system­ atic Branch and Bound search algorithms that exploit partition-based heuristics, BBBT (a new algorithm for which the heuristic information is constructed during search and allows dy­ namic variable/value ordering) and its predeces­ sor BBMB (for which the heuristic information is pre-compiled) and compares them against a number of popular local search...

Neural networks offer an intriguing set of techniques for learning based on the ad-justment of weights of connections between processing units. However, the powerand limitations of connectionist methods for learning, such as the method of backpropagation in parallel distributed processing networks, are not yet entirely clear.We report on a set of experiments that more precisely ...

Scope and Purpose: This paper shows how artificial neural networks can be seamlessly merged with unique evidential reasoning techniques to form a flexible and convenient framework for rep­ resenting and solving complex scheduling problems with many "hard" and "soft" consttaints. The techniques described form the core of SPIKF.. an operational scheduling environment for long­ range scheduling of...

INTRODUCTION Artiicial neural networks constitute a novel class of many-body systems in which the particles are neuron-like units and the interactions are weighted synapse-like connections between these units. 1;2 The most extraordinary feature of these systems is that the interactions are subject to modiication, depending on the states recently visited by the system. Thus, as the network exper...