Model selection for causal theories

Benoît Desjardins

pp. 49-59

in: Maria L. Dalla Chiara, Roberto Giuntini, Federico Laudisa (eds), Language, quantum, music, Berlin, Springer, 1999

Abstract

One of the most central problems in scientific research is the search for explanations to some aspect of nature. This often involves a cycle of data gathering, theorizing, and experimentation. In many scientific fields, including medicine, data comes in the form of statistical distribution information, representing the value of different features for a sample in a population. One of the tasks in research is to discover some structure in that data. In particular, one is interested in finding something about the causal processes explaining the statistical data, in the form of a theory or a model of the aspect of nature under study. Such causal model can then be used as a basis for explanation and experimentation.