Causal Explorer: A Causal Probabilistic Network Learning Toolkit for Biomedical Discovery.
Constantin Aliferis, Ioannis Tsamardinos, Alexander Statnikov, Laura E. Brown · 2003
Causal Probabilistic Networks (CPNs), (a.k.a. Bayesian Networks, or Belief Networks) are well-established representations in biomedical applications such as decision support systems and predictive modeling or mining of causal hypothesis. Increasingly they are also being recognized as a promising formalism for modeling and discovery in bioinformatics and computational biology. Particularly appealing aspects of CPNs are: (a) the existence of welldeveloped theory for induction of causal relationships, and (b) the suitability of these models for creating sound and practical decision support systems. While several public domain and commercial tools exist for modeling and inference with CPNs very few software tools and libraries exist currently that give access to algorithms for CPN induction. To stimulate research with CPNs and enhance the researchers' computing arsenal, we have developed a software library, called Causal Explorer, that implements a suit of global, local and partial CPN induction algorithms. The toolkit emphasizes causal discovery algorithms. Approximately half of the algorithms are enhanced implementations of well established algorithms, and the remaining ones are novel local and partial algorithms that scale to thousands of variables and thus are particularly suitable for modeling in massive datasets. The toolkit can be used to gain insight in the structure of the studied domain, to select promising variables for subsequent experimentation or detailed modeling, or to derive a minimal set of optimal predictors for classification.