LEARNING CAUSAL NETWORKS FROM DATA - A SURVEY AND A NEW ALGORITHM FORRECOVERING POSSIBILISTIC CAUSAL NETWORKS

Causal concepts play a crucial role in many reasoning tasks. Organised as a model revealing the causal structure of a domain, they can guide inference through relevant knowledge. This is an especially difficult kind of knowledge to acquire, so some methods for automating the indu ction of causal models from data have been put forth. Here we review t hose that have a graph representation. Most work has been done on the problem of recovering belief nets from data but some extensions are ap pearing that claim to exhibit a true causal semantics. We will review the analogies between belief networks and ''true'' causal networks and to what extent methods for learning belief networks can be used in le arning causal representations. Some new results in recovering possibil istic causal networks will also be presented.