Interactive Explanation Spaces for Understanding AI Predictions in Cardiovascular Disease Risk

Abstract

Although a plethora of research has been published in the literature, providing both qualitative and quantitative analyses of cardiovascular risk, there remains a need to improve interpretability, explainability, and accuracy in the assessment of cardiovascular disease risk. To achieve this, the present study proposes a methodology that extracts knowledge from data to assess cardiovascular disease risk while also offering both local and global explanations that justify the underlying theory made and why in some cases no definite decision can be taken.

Ähnliche Arbeiten