References & Citations
Statistics > Machine Learning
Title: Gaussian Probabilities and Expectation Propagation
(Submitted on 29 Nov 2011 (v1), last revised 28 Nov 2013 (this version, v2))
Abstract: While Gaussian probability densities are omnipresent in applied mathematics, Gaussian cumulative probabilities are hard to calculate in any but the univariate case. We study the utility of Expectation Propagation (EP) as an approximate integration method for this problem. For rectangular integration regions, the approximation is highly accurate. We also extend the derivations to the more general case of polyhedral integration regions. However, we find that in this polyhedral case, EP's answer, though often accurate, can be almost arbitrarily wrong. We consider these unexpected results empirically and theoretically, both for the problem of Gaussian probabilities and for EP more generally. These results elucidate an interesting and non-obvious feature of EP not yet studied in detail.
Submission history
From: John Cunningham [view email][v1] Tue, 29 Nov 2011 14:59:12 GMT (2328kb,D)
[v2] Thu, 28 Nov 2013 18:58:52 GMT (3632kb,D)
Link back to: arXiv, form interface, contact.