论文部分内容阅读
现代计算机与神经系统有着千丝万缕的关系:在二次世界大战之际,数学家维纳、冯·诺依曼和一批神经生物学家们共同建立了现代计算机原理。在80年代之前,计算机主要是机械地执行算法指令。近几年来,加州理工学院的科学家们提出了光学神经计算机的设想,并在实验室内建立了实验性的模式识别系统,该系统有朝一日会有效地解决随机问题。光学神经计算机的关键在于要按神经元在大脑中排列的方式来布置光学元件。一、如何用计算机来识别“鸟”一般来说,一个牙牙学语的小孩就能轻而易举地从图册中辨别出一只鸟来,但是用一台功能很强的超级计算机来执行这个任务,却不能得到满意的结果。为什么用计算机来辨别物体会如此困难呢?这个问题
Modern computers are inextricably linked to the nervous system: during the Second World War, mathematicians Wiener, von Neumann and a group of neurobiologists worked together to establish modern computer principles. Before the 1980s, computers mainly executed algorithmic instructions mechanically. In recent years, Caltech scientists have come up with the idea of an optical neural computer and have built experimental pattern recognition systems in the lab that will someday effectively solve stochastic problems. The key to an optical nerve computer is to arrange the optical elements in such a way that the neurons line up in the brain. First, how to identify “birds” with a computer In general, a babbling child can easily identify a bird from the album, but use a powerful supercomputer to perform this task, But can not get satisfactory results. Why use a computer to identify objects so difficult? This problem