Fig. 3: Most influential feature interactions in the reinforcement learning algorithm. | npj Digital Medicine

Fig. 3: Most influential feature interactions in the reinforcement learning algorithm.

From: The impact of using reinforcement learning to personalize communication on medication adherence: findings from the REINFORCE trial

Fig. 3

This figure shows the model weights from the feature importance score from the reinforcement learning algorithm, which indicates which features were more or less importance to the model. The weights above were the 20 most influential features, ranked from highest to lowest. Abbreviations: HbA1c, glycated hemoglobin A1c; SGLT2, sodium-glucose cotransporter-2.

Back to article page