Probabilistic networks, which provide compact descriptions of complex stochastic relationships among several random variables, are rapidly becoming the tool of choice for uncertain reasoning in artificial intelligence. We show that networks with fixed structure containing hidden variables can be learned automatically from data using a gradient-descent mechanism similar to that used in neural networks. We also extend the method to networks with intensionally represented distributions, including networks with continuous variables and dynamic probabilistic networks. Because probabilistic networks provide explicit representations of causal structure, human experts can easily contribute prior knowledge to the training process, thereby significantly improving the learning rate. Adaptive probabilistic networks (APNs) may soon compete directly with neural networks as models in computational neuroscience as well as in industrial and financial applications. 1 Introduction Intelligent systems, ...

This paper introduces an information theoretic approach to verification of modular causal probabilistic models. We assume systems which are gradually extended by adding new functional modules, each having a limited domain knowledge captured by a local Bayesiannetwork. Differentmodulesoriginatefromindependent design processes. We assume that the local models are correct, which, however does not guarantee globally coherent inference in composed systems. The introduced method supports discovery of significant inter module dependencies which are ignored in the assembledBayesiannetwork.

Motivated by recent work in phonotactic learning (Hayes and Wilson 2008, Albright 2009), this paper shows how to define feature-based probability distributions whose parameters can be provably efficiently estimated. The main idea is that these distributions are defined as a product of simpler distributions (cf. Ghahramani and Jordan 1997). One advantage of this framework is it draws attention to what is minimally necessary to describe and learn phonological feature interactions in phonotactic patterns. The “bottom-up” approach adopted here is contrasted with the “top-down ” approach in Hayes and Wilson (2008), and it is argued that the bottom-up approach is more analytically transparent. 1