Variational hybridization and transformation for large inaccurate noisy-or networks

TL;DR

This paper introduces a hybrid variational inference method for large, noisy-or Bayesian networks that improves speed, stability, and scalability for real-time medical diagnosis applications.

Contribution

It proposes a novel hybrid inference approach that estimates variational parameters without relying on disease posteriors or priors, enhancing efficiency and recyclability.

Findings

Faster inference with recyclable computation.

Stable transformation ranking despite large prior variances.

Scalable performance demonstrated on large real and synthetic networks.

Abstract

Variational inference provides approximations to the computationally intractable posterior distribution in Bayesian networks. A prominent medical application of noisy-or Bayesian network is to infer potential diseases given observed symptoms. Previous studies focus on approximating a handful of complicated pathological cases using variational transformation. Our goal is to use variational transformation as part of a novel hybridized inference for serving reliable and real time diagnosis at web scale. We propose a hybridized inference that allows variational parameters to be estimated without disease posteriors or priors, making the inference faster and much of its computation recyclable. In addition, we propose a transformation ranking algorithm that is very stable to large variances in network prior probabilities, a common issue that arises in medical applications of Bayesian networks. In experiments, we perform comparative study on a large real life medical network and scalability study on a much larger (36,000x) synthesized network.