Neural nets as models of global socio-political behavior

被引：0

作者：

Kaufman, Jonathan J. ^{[1
]}

机构：

[1] City Univ of New York, United States

来源：

Neural Networks | 1988年 / 1卷 / 1 SUPPL期

关键词：

D O I：

10.1016/0893-6080(88)90474-1

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

It is proposed that neural nets can serve as models of global socio-political behavior, and that this approach can lead to important insights into the organization of society. The dynamical behavior of various types of social systems is then explored. One model of modern society is described with a set { fij } of connection weights which satisfy the following assumption: there are, for the most part, a large number of small, isolated units or groups of people within which the connections (weights) are strong whereas between each group the connections are relatively weak. The dynamical behavior of such a system is shown to exhibit oscillations using computer simulations. A more complex and realistic model of modern society takes into account the inhomogeneity of the net itself. A model is also proposed which offers global stability.

引用