A visual toolbox for modeling and testing multi-net neural systems

作者：Melin Patricia^*; Mendoza Olivia; Castillo Oscar; Ramon Castro Juan

来源：Computer Applications in Engineering Education, 2013, 21(1): 164-184.

DOI：10.1002/cae.20459

• multi-net systems
• modular neural networks
• ensemble neural networks
• image recognition
• time series prediction
• edge detection

Neural networks (NN) and muti-net neural systems are popular and important topics for graduate engineering students. This paper describes a visual toolbox that has been developed to perform experiments with new multi-net neural structures, including ensemble and modular NN approaches. This toolbox can be considered an important contribution for graduate engineering education and research because it is a complete tool that allows to draw models, set parameters, save projects, and generate files with results in a user friendly visual environment. Experimental results with a group of students after using the toolbox show a significant improvement on the acquired knowledge of NNs concepts.

• 出版日期2013-3