* If you want to update the article please login/register
Convolutional Neural Networks are deep neural networks that represent images in image classification, image detection, and image processing are common goals. CNN is the most effective and state-of-the-art technique in image recognition tasks until today. A Neural Architecture Search is a way to produce an accurate network diagram. In this paper, an evolutionary algorithm based on the multiple Sequence Alignment algorithm is suggested to design an appropriate neural network without the burden of building manually. Cartesian Genetic Programming, a new crossover procedure based on the multiple Sequence Alignment algorithm, is suggested by an evolutionary algorithm based on the multiple Sequence Alignment algorithm.
Source link: https://doi.org/10.1007/s11063-022-11093-0
* Please keep in mind that all text is summarized by machine, we do not bear any responsibility, and you should always check original source before taking any actions