Skip to main content
. Author manuscript; available in PMC: 2021 Nov 5.
Published in final edited form as: Int Conf Big Data Smart Comput. 2021 Mar 10;2021:10.1109/bigcomp51126.2021.00021. doi: 10.1109/bigcomp51126.2021.00021

Fig. 3.

Fig. 3.

Real gene expression via RNASeq data are used as features. A. Regression with the cognition score as the outcome; MPRidge employs the least-square loss as the unregularized base learner. B. Binary classification with the clinician’s diagnosis (AD versus non-AD) as the outcome; MPRidge uses the hinge loss as the unregularized base learner. Both real data examples show a near-match in out-of-sample risks, especially at (α*, λ*) which denotes the matched parameter pair minimizing out-of-sample risk.