Model Averaging for Prediction with Discrete Bayesian Networks

@article{Dash2004ModelAF,
title={Model Averaging for Prediction with Discrete Bayesian Networks},
author={Denver Dash and Gregory F. Cooper},
journal={Journal of Machine Learning Research},
year={2004},
volume={5},
pages={1177-1203}
}

Published 2004 in Journal of Machine Learning Research

In this paper 1 we consider the problem of performing Bayesian model-avera ging over a class of discrete Bayesian network structures consistent with a par tial o dering and with bounded in-degree k. We show that forN nodes this class contains in the worst-case at least Ω( (N/2 k )N/2 ) distinct network structures, and yet model averaging over these structu res can be performed using O( (N k ) ·N) operations. Furthermore we show that there exists a single B ay sian network that defines a joint… CONTINUE READING

Pedro Domingos . Bayesian averaging of classifiers and the overfitting problem On the optimality of the simple Bayesian classifier under zero - one loss