P32 - Visualizing Score Contributions in BBQ-Trees
- Event
- iCCC2026 - iCampus Cottbus Conference
2026-05-05 - 2026-05-07
Cottbus - Band
- Poster
- Chapter
- Kommunikation
- Author(s)
- A. Stahl - BTU Cottbus-Senftenberg, Cottbus
- Pages
- 248 - 251
- DOI
- 10.5162/iCCC2026/P32
- ISBN
- 978-3-910600-10-2
- Price
- free
Abstract
Interpretable decisionmaking is essential in medical and health-related applications. BBQ-Trees (BBQTs) combine decision-tree structure with quantum-inspired logic, with advantages to accuracy and interpretability. However, their evaluation scheme can make per-sample explanations challenging. We propose a visualization method that maps subtree contributions to colors in order to reveal which decision paths drive an individual prediction. Using publicly available medical datasets as illustrative examples, we show that the method makes BBQT predictions more transparent without altering their predictive behavior.