生物神经网络仿真中数据表达和并行处理

Data Representation and Parallel Processing in Computer Simulation for Biological Neural Networks

生物神经网络可以处理信息和产生特定的电特性,理解其信息产生和传递机理,对生物学将具有重要的意义。由于神经网络内部复杂的非线性反馈,进行各种可能的实验十分困难。为此,将信息学与生物学相结合,数学模型与实验数据相结合,建立了神经网络计算机仿真系统。在该仿真系统中,提出了面向对象的数据表达,并实现了并行处理,从而极大地提高了仿真系统的效率。该仿真系统已在许多国家科研和教学机构得到广泛使用。

Biological neural networks (BNN) have the ability to process information and generate a specific pattern of electrical activity. In the biology it is significant to understand the mechanism of generating and transmitting the information. As there exists complicated nonlinear feedbacks inside the neural networks, it is very difficult to make all possible experiments on it. Through integrating information science into biology, combining mathematical models with experiment data, one computer simulation system was established. In the simulation system, object-oriented data representation was presented, and parallel processing was implemented, which greatly improved its efficiency. It has been widely applied in research institutes and universities of many countries.