Abstract

Neural network algorithms have impressively demonstrated the capability of modeling spatial information. On the other hand, the application of parallel distributed models to processing of temporal data has been severely restricted. This work introduces a novel technique that adds the dimension of time to the well-known backpropagation neural network algorithm. The paper cites several reasons why the inclusion of automated spatial and temporal associations are crucial to effective systems modeling. An overview of other works that also model spatiotemporal dynamics is furnished. In addition, a detailed description of the processes necessary to implement the space-time neural network (STNN) algorithm is provided. The reader is given several demonstrations that illustrate the capabilities and performance of this new architecture.

Author Keywords: neural network algorithms; space-time neural network; backpropagation

Article Outline

• References

Corresponding author. Address correspondence to James A. Villarreal, Information Systems Directorate, Lyndon B. Johnson Space Center, Houston, TX.