J.H.M. Lee and V.W.L. Tam. A Framework for Integrating Artificial Neural Networks and Logic Programming International Journal on Artificial Intelligence Tools 4(1&2), 3--32, June, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....file) 6.1 Hard Graph Coloring Problem To compare the LSDL implementation of GENET versus other GENET implementations and other methods we investigate its performance on a set of hard graph coloring problems. Table 1 shows the experimental results for GENET as described in [DTWZ94] PROCLANN [LT95] an incremental formulation of GENET, DLM [SW98] and GSAT [SLM92] We omit the lazy versions since there is no arc inconsistency in the problems and the performance is similar to that of the non lazy versions. The timing results of GENET represent the median of 10 runs collected on a SUN Sparc ....

J. H. M. Lee and V. W. L. Tam. A framework for integrating artificial neural networks and logic programming. International Journal of Artificial Intelligence Tools, 4(1&2):3-- 32, 1995.

....non binary, non linear and symbolic constraints in a homogeneous and flexible fashion. We have implemented an E GENET simulator for singleprocessor systems and benchmarked our implementation on CSP s of various types. E GENET compares favorably against GENET [16] CHIP [8] and PROCLANN [21], a constraint logic programminglanguage based on GENET. The rest of this paper is organized as follows. In section 2, we explain the deficiency of GENET in handling non binary constraints, followed by an exposition of the E GENET architecture and convergence procedure. Benchmarking results of ....

....problem, and the Hamiltonian path problem. We compare our result with that of CHIP version 4.0.1 [22] which uses traditional constraint propagation and backtracking tree search for constraint solving. Since we do not have access to a GENET implementation, we compare instead to PROCLANN [21], which is a constraint logic programming language with GENET as the constraint solver. Wherever possible, we quote results of GENET from the paper of Davenport et al. 16] timing of which is obtained on a SUN SPARCclassic. All benchmarking is performed on a SUN SPARCstation 10 model 30. Timing ....

J.H.M. Lee and V.W.L. Tam, "A framework for integrating artificial neural networks and logic programming", International Journal on Artificial Intelligence Tools, 1995, (to appear).

....than propagation based solvers on large or hard instance of CSP s. The problem we examine is how such solvers can be used efficiently in CLP systems. Earlier approaches that considered incorporating stochastic solvers into CLP systems are restricted to one of the models we discuss. Lee and Tam [9] required the program to execute only a single derivation so that backtracking could never occur. At the end of the single derivation the stochastic solver (the GENet al..gorithm discussed later) determines a solution. Illera and Ortiz [6] use a genetic algorithm to search for a good solution to a ....

J.H.M. Lee and V.W.L. Tam. A Framework for Integrating Artificial Neural Networks and Logic Programming International Journal on Artificial Intelligence Tools 4(1&2), 3--32, June, 1995.

....than propagation based solvers on large or hard instance of CSP s. The problem we examine is how such solvers can be used efficiently in CLP systems. Earlier approaches that considered incorporating stochastic solvers into CLP systems are restricted to one of the models we discuss. Lee and Tam [13] required the program to execute only a single derivation so that backtracking could never occur. At the end of the single derivation the stochastic solver (the GENet al..gorithm discussed later) determines a solution. Both Illera and Ortiz [7] and Kok et al. 11] use genetic algorithms to search ....

J.H.M. Lee and V.W.L. Tam. A Framework for Integrating Artificial Neural Networks and Logic Programming International Journal on Artificial Intelligence Tools 4(1&2), 3--32, June, 1995.

No context found.

J.H.M. Lee and V.W.L. Tam. A Framework for Integrating Artificial Neural Networks and Logic Programming International Journal on Artificial Intelligence Tools 4(1&2), 3--32, June, 1995.

....kinds of solvers can be considerably more efficient than propagation based solvers on large or hard instance of CSP s. The problem we examine is how can such solvers be used efficiently in CLP systems. Two earlier approaches considered incorporating stochastic solvers into CLP systems. The first [7, 8] required the program to execute only a single derivation so that backtracking could never occur. The second [6] used the stochastic solver in combination with a propagation based solver. The propagation based solver directed the derivation, and the stochastic solver was invoked at the end of a ....

J.H.M. Lee and V.W.L. Tam. A Framework for Integrating Artificial Neural Networks and Logic Programming International Journal on Artificial Intelligence Tools 4(1&2), pages 3--32, June, 1995.

....to extract a solution. An attractive proposal is to incorporate stochastic methods into CLP, yielding an efficient and expressive constraint programming language. Two earlier approaches consider replacing the usual propagation based solver in CHIP like CLP languages by stochastic solvers. PROCLANN [11] uses a stochastic solver based on artificial neural networks. The language allows only a single thread of execution so that backtracking never occurs. Constraints are collected during derivation and the stochastic solver is exercised at the end of a derivation as a solution finder. PROCLANN, ....

J.H.M. Lee and V.W.L. Tam. A framework for integrating artificial neural networks and logic programming. International Journal on Artificial Intelligence Tools, 4(1&2):3--32, June 1995.

....to extract a solution. An attractive proposal is to incorporate stochastic methods into CLP, yielding an efficient and expressive constraint programming language. Two earlier approaches consider replacing the usual propagation based solver in CHIP like CLP languages by stochastic solvers. PROCLANN [8] uses a stochastic solver based on artificial neural networks. The language allows only a single thread of execution so that backtracking never occurs. Constraints are collected during derivation and the stochastic solver is exercised at the end of a derivation as a solution finder. PROCLANN, ....

J.H.M. Lee and V.W.L. Tam. A framework for integrating artificial neural networks and logic programming. International Journal on Artificial Intelligence Tools, 4(1&2):3--32, June 1995.

....than propagation based solvers on large or hard instances of CSPs. The problem we examine is how such solvers can be used e#ciently in CLP systems. Earlier approaches that considered incorporating stochastic solvers into CLP systems were restricted to one of the models we discuss. Lee and Tam [LT95] required the program to execute only a single derivation, so that backtracking could never occur. At the end of the single derivation, the stochastic solver (the GENet al..gorithm discussed later) determines a solution. Both Illera and Ortiz [IO95] and Kok et al. KMMR96] used genetic algorithms to ....

Jimmy H. M. Lee and Vincent Tam. A framework for integrating artificial neural networks and logic programming.<F4.569e+05> International Journal on Artificial Intelligence<F5.341e+05> Tools, 4(1--2):3--32, June 1995.

No context found.

Lee, J.H.M. & Tam, V.W.L., A framework for integrating artificial neural networks and logic programming, International Journal on Artificial Intelligence Tools, Vol.4, Nos.1&2, June 1995, 3-32

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC