A nonparametric Bayesian method of translating machine learning scores to probabilities in clinical decision support

Connolly B., Cohen K. B., Santel D., Bayram U., Pestian J.

BMC BIOINFORMATICS, vol.18, 2017 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 18
  • Publication Date: 2017
  • Doi Number: 10.1186/s12859-017-1736-3
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Keywords: Statistics, Nonparametric, Bayesian, Calibration, Machine learning, PATIENT OUTCOMES, PRACTITIONER PERFORMANCE, CARDIOVASCULAR-DISEASE, SYSTEMS, DIAGNOSIS, RISK
  • Çanakkale Onsekiz Mart University Affiliated: No


Background: Probabilistic assessments of clinical care are essential for quality care. Yet, machine learning, which supports this care process has been limited to categorical results. To maximize its usefulness, it is important to find novel approaches that calibrate the ML output with a likelihood scale. Current state-of-the-art calibration methods are generally accurate and applicable to many ML models, but improved granularity and accuracy of such methods would increase the information available for clinical decision making.