Towards measures of intelligence based on semiotic control.

Joslyn, C.

Technical Information Center Oak Ridge Tennessee

关键词: Computerized control systems;    Learning;    Training;    Measuring methods;    Systems analysis;   

RP-ID  :  DE2001768903

学科分类：工程和技术（综合）

美国|英语

来源: National Technical Reports Library

【 摘 要 】

We address the question of how to identify and measure the degree of intelligence in systems. We define the presence of intelligence as equivalent to the presence of a control relation. We contrast the distinct atomic semiotic definitions of models and controls, and discuss hierarchical and anticipatory control. We conclude with a suggestion about moving towards quantitative measures of the degree of such control in systems.