Skip to main content
Fig. 3 | BMC Medical Informatics and Decision Making

Fig. 3

From: A qualitative research framework for the design of user-centered displays of explanations for machine learning model predictions in healthcare

Fig. 3

Prototypes of explanation displays utilizing different design options. Design options used in each prototype are listed as follows: a) unit of explanation, b) organization of explanation units, c) dimensionality, d) risk representation, and e) explanation display format. Prototype 1 options: a) individual features, b) none, c) interactive explanation size, static explanation unit granularity, d) odds, e) tornado plot. Prototype 2 options: a) feature groups, b) influence groups, c) interactive explanation size, interactive explanation unit granularity, d) probability, e) tornado plot. Prototype 3 options: a) feature groups, b) influence groups, c) static explanation size, static explanation unit granularity, d) probability, e) force plot. Prototype 4 options: a) individual features, b) influence groups, c) interactive explanation size, static explanation unit granularity, d) probability, e) tornado plot. Prototype 5 options: a) feature groups, b) influence groups and assessment groups, c) interactive explanation size, interactive explanation unit granularity, d) probability, e) tornado plot

Back to article page