2008 | OriginalPaper | Chapter
Partitioning for High Performance of Predicting Dynamical Behavior of Color Diffusion in Water using 2-D tightly Coupled Neural Cellular Network
Authors : A. Suratanee, K. Na Nakornphanom, K. Plaimas, C. Lursinsap
Published in: Modeling, Simulation and Optimization of Complex Processes
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
The 2-D tightly coupled neural cellular network is developed to simulate the diffusion characteristics of colored liquid dropped onto water surface by generating its own diffusion image. To learn the diffusion characteristics, a large training set was divided into data subsets by extracting the significant feature patterns of the diffusion and simultaneously training the neural cellular network individually. Using this technique for reducing the training-time, increasing the performance, and facilitating the recognition of large data sets, many sub optimal neural networks were developed to replace of one network. Additionally, the result of the partitioning data achieved the speedup of 17.9 times for 12 networks and 605,267 data patterns. The accuracy of the simulated behaviour is more than 90 percent compared with the actual event.