IEEE P3347

New IEEE Standard - Active - Draft.
This standard defines a text description method for neural networks. Data requirements, basic unit in neural networks, schemas for neural network architectures, and application scenarios are also discussed. Institutions, enterprises and researchers that describe neural network architecture follow the general description method outlined here. With this standard, newly generated neural network architectures could be easily shared and exchanged. This would make neural network algorithms very easily transferred because they will not depend on the particular neural network platform or libraries.

This standard defines a unified text approach to describing neural network architectures. This standard specifies data requirements, coding formats of neurons in neural networks, schemas for neural network architectures, and application scenarios.
This standard defines a text description method for neural networks that can be used for data mining. This leads to new state of arts that researchers never explored, which in turn enhances and enlarges the application scenario of neural network models. Newly generated neural network architectures are easily shared and exchanged via this standard. This makes neural network algorithms very easily transferred because they do not depend on the particular neural network platform or libraries.