Computational Neural Science on the Connection Machine: Adaptative Map Algorithms for Real World Problems

计算神经科学研究大脑的信息处理，并采用神经并行计算策略进行数字信息处理，其一个重要目标是缩小大脑中神经元和突触的高密度与当今计算模型中的低密度之间的差距。在这方面，连接机，一个每个处理器具有大存储容量的大型并行计算机，提供了一个理想的研究工具。研究人员将利用这台新机器来缩小当前大脑模型与现实之间的差距。这项工作本质上是高度计算性的，但它将汇集来自不同学科的概念和技术，如神经生物学、计算机科学和计算物理学。

One important aim of Computational Neural Science which studies information processing brains and adopts neural parallel computational strategies to digital information processing, is to narrow the gap between the high density of neurons and synapses found in brains and the low density in today's computational models. In this respect the Connection Machine, a massive parallel computer with a large storage capacity per processor, provides an ideal research tool. The researchers will exploit this new machine to narrow the gap between current models of the brain and reality. This work will be highly computational in nature, but it will bring together concepts and techniques from disparate disciplines, names Neurobiology, Computer Science, and Computational Physics.