Self-organizing feature maps and the travelling salesman problem

Angéniol, Bernard; Vaubois, Gaël de La Croix; Texier, Jean-Yves Le

doi:10.1016/0893-6080(88)90002-0

Cited by 326 publications

(125 citation statements)

References 6 publications

Supporting

Mentioning

106

Contrasting

Unclassified

Order By: Relevance

“…Angniol et al [37] had done extensive analysis on the number of neurons in SOM. To avoid oscillation of neurons between different cities, they proposed that the number of neurons should be greater than number of cities (M ≥ 3N ).…”

Section: The Kohonen Network Incorporating Explicit Statisticsmentioning

confidence: 99%

“…To avoid oscillation of neurons between different cities, they proposed that the number of neurons should be greater than number of cities (M ≥ 3N ). In our study we assume a fixed number of neurons (M = 5N ) but a variable number of neurons was also studied by Angniol et al [37] and Boeres et al [39]. …”

Section: The Kohonen Network Incorporating Explicit Statisticsmentioning

confidence: 99%

“…2. KNIES [24] has shown better performance than the Pure Kohonen network, the Guilty Net of Burke and Damany [36], and the approach of Angniol et al [37]. 3.…”

Section: The Kohonen Network Incorporating Explicit Statisticsmentioning

confidence: 99%

See 2 more Smart Citations

Parallelized neural network system for solving Euclidean traveling salesman problem

Avşar

Aliabadi

2015

Applied Soft Computing

View full text Add to dashboard Cite

Section: The Kohonen Network Incorporating Explicit Statisticsmentioning

confidence: 99%

Section: The Kohonen Network Incorporating Explicit Statisticsmentioning

confidence: 99%

See 1 more Smart Citation

Parallelized neural network system for solving Euclidean traveling salesman problem

Avşar

Aliabadi

2015

Applied Soft Computing

View full text Add to dashboard Cite

“…L'algorithme de Bonomi et Lutton [2], optimisation de celui de Kirkpatrick, est le meilleur des algorithmes connus de l'auteur [1,3,4,5]. De même que l'algorithme de relaxation isothermique, il ne considère que des échanges de couples (2-OPT) d'arcs voisins.…”

Section: Comparaison Avec Les Méthodes Connexionistesunclassified

Des opérateurs d'échange et une méthode de relaxation pour le problème du voyageur de commerce

Schnetzler¹

1992

RAIRO-Oper. Res.

View full text Add to dashboard Cite

“…There are many types of artificial neural networks and more details can be found in Potvin (1993). The neural-network scheme, first developed by Angeniol et al (1988) is derived from the Kohonen's Self-Organising Map (SOM). An advantage of the SOM is that the resulting map is automatically ordered in the data space.…”

Section: Methodsmentioning

confidence: 99%

Application of high-resolution borehole image analysis in conjunction with wellbore geophysical log data to improve coal characterisation in Permian coal measures, Bowen Basin, Australia

Roslin¹

View full text Add to dashboard Cite

Coal reservoirs are unconventional in the sense that the scale of vertical and lateral variability is high when compared to conventional sandstone reservoirs. This requires higher resolution data (metres to decimetres to sub-microns levels) for calculation of reservoir properties. Rock core is the best, but it's expensive to obtain and destroyed during analysis. Recent advances in geophysical imaging of the wellbore provide "pseudo-core" from which to characterise the reservoir, but require advanced processing techniques coupled with ground-truthing to make them a reliable alternative.This project sought to develop methods for advanced coal seam gas reservoir characterisation using wellbore geophysical data and image logs, and to use this information to interpret coal seam gas reservoir properties, in particular coal lithotype, in the context of geological development of the Bowen Basin.The first problem solved during the research was a conversion of borehole electrical images into a microresistivity curve. This procedure allowed the researcher to make a step from qualitative analysis of borehole images to the quantitative one. The resulting microresistivity curve was added to the wireline suite for geostatistical cluster analysis and the development of electrofacies. Thus, the resulting wireline suite consisted of gamma ray, bulk density, photo-electric factor, image-based microresistivity and deep laterolog wireline data.Electrofacies analysis produced a coal lithotype profile which was validated by the results of other studies such as mm scale brightness profile logging and maceral content analysis. Implementation of the microresistivity curve method enhanced the resolution of coal profiling to decimetres level, while photo-electric curve allowed distinguishing between vitrinite-and inertinite-rich low density coal.The results of electrofacies analysis were obtained for 26 wells spread over the Northern Bowen Basin area and the distribution of the coal electrofacies were analysed for both stratigraphical and geographical trends. It was observed that the inertinite-rich coal interpreted from electrofacies analysis increases towards the top of Late Permian coal and towards the north of the Northern Bowen Basin area. These results were consistent with previously observed trends defined by petrographic analysis. 3 Overall, this research project has resulted in the development of a new method for coal characterisation which goes beyond simple cut off values for density, and allows coal lithotype advanced characterisation based on high-resolution wireline borehole data.This method was tested in a case study where stratigraphical and geographical distributions of coal lithotypes were generated over the Northern Bowen Basin area.4

show abstract

Self-organizing feature maps and the travelling salesman problem

Cited by 326 publications

References 6 publications

Parallelized neural network system for solving Euclidean traveling salesman problem

Parallelized neural network system for solving Euclidean traveling salesman problem

Des opérateurs d'échange et une méthode de relaxation pour le problème du voyageur de commerce

Application of high-resolution borehole image analysis in conjunction with wellbore geophysical log data to improve coal characterisation in Permian coal measures, Bowen Basin, Australia

Contact Info

Product

Resources

About