Using Deep Neural Network Approximate Bayesian Network

TL;DR

This paper introduces a deep neural network approach to efficiently approximate Bayesian network posterior probabilities, outperforming traditional methods in speed and accuracy, especially for medium-sized networks, with easy GPU parallelization.

Contribution

It demonstrates that deep neural networks can effectively learn and approximate the joint probability distribution of Bayesian networks, offering a faster and more scalable alternative to traditional sampling methods.

Findings

Our method is faster than likelihood weighting sampling.

It achieves higher accuracy in medium-sized Bayesian networks.

The model saturates with fewer training examples, needing less data for good performance.

Abstract

We present a new method to approximate posterior probabilities of Bayesian Network using Deep Neural Network. Experiment results on several public Bayesian Network datasets shows that Deep Neural Network is capable of learning joint probability distri- bution of Bayesian Network by learning from a few observation and posterior probability distribution pairs with high accuracy. Compared with traditional approximate method likelihood weighting sampling algorithm, our method is much faster and gains higher accuracy in medium sized Bayesian Network. Another advantage of our method is that our method can be parallelled much easier in GPU without extra effort. We also ex- plored the connection between the accuracy of our model and the number of training examples. The result shows that our model saturate as the number of training examples grow and we don't need many training examples to get reasonably good result. Another contribution of our work is that we have shown discriminative model like Deep Neural Network can approximate generative model like Bayesian Network.