SHAP Algorithm for Healthcare Data Classification

• Published in: Hybrid Artificial Intelligent Systems Vol. 13469; pp. 363 - 374
• Main Authors: Mihirette, Samson, Tan, Qing
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2022; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Data privacy; Data privacy broker; Healthcare Information System (HIS); SHAP
• ISBN: 9783031154706; 3031154703
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-031-15471-3_31

To strengthen the healthcare data privacy protecting techniques and ensure the transparency of healthcare data exchange, many data privacy-preserving methods have been introduced. This paper highlights privacy concerns and introduces techniques and research directions towards data privacy in Healthcare Information Systems (HIS). The paper demonstrates the use and the power of the Shapley Additive exPlanations (SHAP) algorithm to identify and classify critical data elements that can put personal privacy at risk within a dataset. A conceptual patient-centric healthcare information system architecture with a data broker is proposed in this paper. The proposed architecture also includes the privacy broker that leverages application programming interface services and integration middleware in safeguarding healthcare data privacy.