Basic Study
Copyright ©The Author(s) 2018. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Clin Oncol. Sep 14, 2018; 9(5): 98-109
Published online Sep 14, 2018. doi: 10.5306/wjco.v9.i5.98
Tunable structure priors for Bayesian rule learning for knowledge integrated biomarker discovery
Jeya Balaji Balasubramanian, Vanathi Gopalakrishnan
Jeya Balaji Balasubramanian, Intelligent Systems Program, School of Computing and Information, University of Pittsburgh, Pittsburgh, PA 15260, United States
Vanathi Gopalakrishnan, Department of Biomedical Informatics, School of Medicine, University of Pittsburgh, Pittsburgh, PA 15206, United States
Author contributions: Balasubramanian JB developed the concept, conducted the research, and prepared the first draft of the manuscript in consultation with research mentor and senior author Gopalakrishnan V; All authors contributed to writing and editing the manuscript.
Supported by National Institute of General Medical Sciences of the National Institutes of Health, No. R01GM100387.
Conflict-of-interest statement: The authors declare no conflicts of interest with respect to the submitted manuscript.
Open-Access: This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Correspondence to: Vanathi Gopalakrishnan, PhD, Associate Professor, Department of Biomedical Informatics, School of Medicine, University of Pittsburgh, Room 530, 5607 Baum Boulevard, Pittsburgh, PA 15206, United States. vanathi@pitt.edu
Telephone: +1-412-6243290 Fax: +1-412-6245310
Received: April 27, 2018
Peer-review started: April 27, 2018
First decision: July 9, 2018
Revised: July 24, 2018
Accepted: August 5, 2018
Article in press: August 5, 2018
Published online: September 14, 2018
Processing time: 140 Days and 15.9 Hours
Core Tip

Core tip: Bayesian rule learning is a unique rule learning algorithm that infers rule models from searched Bayesian networks. We extended it to allow the incorporation of prior domain knowledge using a mathematically robust Bayesian framework with structure priors. The hyperparameter of the structure priors enables the user to control the influence of their specified prior knowledge. This opens up many possibilities including incorporating uncertain knowledge that can interact with data accordingly during inference.