Bayesian networks for interpretable machine learning and optimization

Bojan Mihaljević

doi:10.1016/j.neucom.2021.01.138

2021-10-01

Bojan Mihaljević *

Concha Bielza

Pedro Larrañaga

DOI：10.1016/j.neucom.2021.01.138

原文链接

查看原文

摘要

En 中文

As artificial intelligence is being increasingly used for high-stakes applications, it is becoming more and more important that the models used be interpretable. Bayesian networks offer a paradigm for inter-pretable artificial intelligence that is based on probability theory. They provide a semantics that enables a compact, declarative representation of a joint probability distribution over the variables of a domain by leveraging the conditional independencies among them. The representation consists of a directed acyclic graph that encodes the conditional independencies among the variables and a set of parameters that encodes conditional distributions. This representation has provided a basis for the development of algo-rithms for probabilistic reasoning (inference) and for learning probability distributions from data. Bayesian networks are used for a wide range of tasks in machine learning, including clustering, super -vised classification, multi-dimensional supervised classification, anomaly detection, and temporal mod-eling. They also provide a basis for estimation of distribution algorithms, a class of evolutionary algorithms for heuristic optimization. We illustrate the use of Bayesian networks for interpretable machine learning and optimization by presenting applications in neuroscience, the industry, and bioin-formatics, covering a wide range of machine learning and optimization tasks. (c) 2021 Published by Elsevier B.V.

Keyword:

Interpretability

Explainable machine learning

Probabilistic graphical models