Generalized Ideal Parent (GIP): Discovering non-Gaussian Hidden Variables

A formidable challenge in uncertainty modeling in general, and when learning Bayesian networks in particular, is the discovery of unknown hidden variables. Few works that tackle this task are typically limited to discrete or Gaussian domains, or to tree structures. We propose a novel approach for discovering hidden variables in flexible nonGaussian domains using the powerful class of Gaussian copula networks. Briefly, we define the concept of a hypothetically optimal predictor of variable, and show how it can be used to discover useful hidden variables in the expressive framework of copula networks. We demonstrate the merit of our approach for learning succinct models that generalize well in several real-life domains.