Cyclic Causal Complexes

Abstract

Causal commonsense reasoning perceptions play an essential role in human decision-making. A known cause/effect relationship has a high decision value. Knowledge of at least some relationships is inherently imprecise. Causal complexes are groupings of smaller causal relations that can make up a larger grained causal object. Usually, commonsense reasoning is more successful in reasoning about a few large-grained events than many finer-grained events. However, larger-grained causal objects are necessarily more imprecise. A satisficing solution might be to develop large-grained solutions and then develop finer-grain objects when the impreciseness of the larger-grain is unsatisfactory. Often, a causal relationship is represented by a network with conditioned edges (probability, possibility, randomness, etc.). Various kinds of representational graphs and models can be used. One class of needed necessary descriptions are cycles, including mutual causal dependencies, both with non-cumulative effects and cumulative effects (including feedback). Without cyclic descriptions, there will be an incomplete representation of the variety and wealth of causal constructions used in science as well as in everyday life. Causal Bayes networks have received significant attention; a significant weakness is that they do not allow cycles; they have other significant restrictions, including independence conditions that include Markoff conditions. This paper discusses general and Bayes causal networks and introduces general imprecise graphic causal models.