Postdoc: Automatic Programming of Decentralized Parallel Architectures

Mitchell 9503162这个博士后研究将开发新技术，用于分散式并行架构（如元胞自动机）的自动编程。 初步研究表明，遗传算法（GAs）是设计一维元胞自动机进行计算的有前途的技术。 主要工作将是将工作从一维系统扩展到二维系统，这将使设计技术能够应用于现实世界的并行计算问题，例如图像处理、模式识别和图像压缩。 这种扩展将需要在空间并行计算理论以及遗传算法的理论和应用方面取得一些进展。

Mitchell 9503162 This postdoctoral research will be on developing novel techniques for the automatic programming of decentralized parallel architectures such as cellular automata. Preliminary research performed has shown that genetic algorithms (GAs) are promising techniques for designing one-dimensional cellular automata to perform computations. The major effort will be in extending the work from one- to two-dimensional systems, which will allow the design techniques to apply to real-world parallel computation problems such as image processing, pattern recognition, and image compression. This extension will require a number of advances in the theory of spatial parallel computation as well as in the theory and application of genetic algorithms.