paperKB
coga / coga-kb
Help
Sign in

Chunk #23 — Results — Discriminative features — Medication

Source
Predicting alcohol use disorder remission: a longitudinal multimodal multi-featured machine learning approach.
Embedded
yes

Text

Adding medications to the PRS models significantly increased model accuracy for EA males—sleep medication, AA males—other medication. (p(PRS vs PRS & sleep medication) <0.001, p(PRS vs PRS & other medication) <0.001, respectively). Depression medication increased the accuracy score of the PRS & EEG-FC model but this increase did not reach significance (p(PRS & EEG vs PRS & EEG & depression medication) = 0.5). In those cases, taking medication predicted maintenance of the AUD state (Table 2 for weight ranking).