Abstract. Probabilistic classification requires the computation of the posterior probability distribution of a class given a data observation. In order to generate posterior samples, an analogy has recently been established between the Bayesian updating problem and the engineering reliability problem which allows reliability methods to be applied to the former. The modification of the BUS (Bayesian Updating with Structural Reliability Methods) formulation is based on the conventional rejection principle and suggests the application of Subset Simulation (SuS) from reliability engineering to sample from posterior distributions. Under the original BUS framework a likelihood multiplier is required to be calculated before the implementation of SuS. A recently proposed algorithm learns the likelihood multiplier automatically. This research proposes the utilization of BUS for Gaussian Process classification. The above framework is illustrated using a benchmark Machine Learning dataset from an engineering application.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.