Construction of nonparametric Bayesian models from parametric Bayes equations

We consider the general problem of constructing nonparametric Bayesian models on infinite-dimensional random objects, such as functions, infinite graphs or infinite permutations. The problem has generated much interest in machine learning, where it is treated heuristically, but has not been studied in full generality in non-parametric Bayesian statistics, which tends to focus on models over probability distributions. Our approach applies a standard tool of stochastic process theory, the construction of stochastic processes from their finite-dimensional marginal distributions. The main contribution of the paper is a generalization of the classic Kolmogorov extension theorem to conditional probabilities. This extension allows a rigorous construction of nonparametric Bayesian models from systems of finite-dimensional, parametric Bayes equations. Using this approach, we show (i) how existence of a conjugate posterior for the nonparametric model can be guaranteed by choosing conjugate finite-dimensional models in the construction, (ii) how the mapping to the posterior parameters of the nonparametric model can be explicitly determined, and (iii) that the construction of conjugate models in essence requires the finite-dimensional models to be in the exponential family. As an application of our constructive framework, we derive a model on infinite permutations, the nonparametric Bayesian analogue of a model recently proposed for the analysis of rank data.

Numerical simulation of post-flame oxidation of hydrocarbons in spark ignition engines

About 50-90 percent of the hydrocarbons that escape combustion during flame passage in spark-ignition engine operation are oxidized in the cylinder before leaving the system. The process involves the transport of unreacted fuel from cold walls towards the hotter burned gas regions and subsequent reaction. In order to understand controlling factors in the process, a transient one-dimensional reactive-diffusive model has been formulated for simulating the oxidation processes taking place in the reactive layer between hot burned gases and cold unreacted air/fuel mixture, with initial and boundary conditions provided by the emergence of hydrocarbons from the piston top land crevice. Energy and species conservation equations are solved for the entire process, using a detailed chemical kinetic mechanism for propane. Simulation results show that the post-flame oxidation process takes place within a reactive layer where intermediate hydrocarbon products are formed at temperatures above 1100-1200 K, followed by a carbon monoxide conversion region closer to the hot burned gases. Model results show that most of hydrocarbons leaving the crevice are completely oxidized inside the cylinder. The largest contribution of remaining hydrocarbons are those leaving the crevice at temperatures below 1400 K. The largest fraction of non-fuel (intermediate) hydrocarbons results from hydrocarbons leaving the crevice when core temperatures are around 1400 K Copyright © 1997 Society of Automotive Engineers, Inc.