Causality and model abstraction
From Semantic Portal Wiki
{{#vardefine:category|Publication}}{{#vardefine:templatename|i.publication}}{{#vardefine:package|smwbp_instance_templates}}
| Edit |
Reference: {{#vardefine:pagename|causality and model abstraction }}
- [[]]
bibtex
{{#vardefine:pagename|Causality and model abstraction }}{{#vardefine:key| }}
abstract: This paper describes a computational approach, based on the theory of causalordering, for inferring causality from an acausal, formal description of aphenomena. Causal ordering, first proposed by Simon, in an asymmetricrelation among the variables in a self-contained equilibrium or dynamic model,which reflects people's intuitive notion of causal dependency relations amongvariables. We extended the theory to cover models consisting of a mixture ofdynamic and equilibrium equations. When people's intuitive causalunderstanding of a situation is based on a dynamic description, the causalordering produced by the extension reflects such intuitive understandingbetter than that of an equilibrium description. As the number of variables ina system increases and the system becomes more complex, model abstractionbecomes essential in reasoning about its behavior. Aggregation of a nearlydecomposable dynamic systems is an abstraction technique that provides aformal justification for commonsense abstraction whose application is easilyobservable in everyday life. The paper examines the close relation betweenaggregation and causal ordering.
download:
- paper:
- slides:
| Abstract | This paper describes a computational appro … This paper describes a computational approach, based on the theory of causalordering, for inferring causality from an acausal, formal description of aphenomena. Causal ordering, first proposed by Simon, in an asymmetricrelation among the variables in a self-contained equilibrium or dynamic model,which reflects people's intuitive notion of causal dependency relations amongvariables. We extended the theory to cover models consisting of a mixture ofdynamic and equilibrium equations. When people's intuitive causalunderstanding of a situation is based on a dynamic description, the causalordering produced by the extension reflects such intuitive understandingbetter than that of an equilibrium description. As the number of variables ina system increases and the system becomes more complex, model abstractionbecomes essential in reasoning about its behavior. Aggregation of a nearlydecomposable dynamic systems is an abstraction technique that provides aformal justification for commonsense abstraction whose application is easilyobservable in everyday life. The paper examines the close relation betweenaggregation and causal ordering. on betweenaggregation and causal ordering. |
| Author | Yumi Iwasaki + |
| Bibtype | techreport + |
| Institution | Knowledge Systems, AI Laboratory + |
| Key | KSL-89-80 + |
| Month | March + |
| Note | Revised March 1993. + |
| Number | KSL-89-80 + |
| Tag | Computer science + |
| Title | Causality and Model Abstraction + |
| Tr id | KSL-89-80 + |
| Year | 1989 + |
