Image Representations of Numerical Simulations for Training Neural Networks

A large amount of data can partly assure good fitting quality for the trained neural networks. When quantity experimental or on-site monitoring is commonly insufficient and difficult to control in engineering practice, numerical simulations provide a controlled high data. Once networks are by such data, they be used predicting properties/responses objects instantly, saving further computing efforts simulation tools. Correspondingly, strategy efficiently transferring input output obtained desirable engineers programmers. In this work, we proposed simple image representation simulations, where all represented images. The temporal spatial information kept greatly compressed. addition, results readable not only computers but also human resources. Some examples given, indicating effectiveness strategy.