TY - JOUR
T1 - Adversarially-Regularized Mixed Effects Deep Learning (ARMED) Models Improve Interpretability, Performance, and Generalization on Clustered (non-iid) Data
AU - Nguyen, Kevin P.
AU - Treacher, Alex H.
AU - Montillo, Albert A.
N1 - Funding Information:
This work was supported in part by Lyda Hill Foundation and in pat by the National Institute Of GeneralMedical Sciences of the National Institutes of Health under Award Number R01GM144486.
Publisher Copyright:
© 1979-2012 IEEE.
PY - 2023/7/1
Y1 - 2023/7/1
N2 - Natural science datasets frequently violate assumptions of independence. Samples may be clustered (e.g., by study site, subject, or experimental batch), leading to spurious associations, poor model fitting, and confounded analyses. While largely unaddressed in deep learning, this problem has been handled in the statistics community through mixed effects models, which separate cluster-invariant fixed effects from cluster-specific random effects. We propose a general-purpose framework for Adversarially-Regularized Mixed Effects Deep learning (ARMED) models through non-intrusive additions to existing neural networks: 1) an adversarial classifier constraining the original model to learn only cluster-invariant features, 2) a random effects subnetwork capturing cluster-specific features, and 3) an approach to apply random effects to clusters unseen during training. We apply ARMED to dense, convolutional, and autoencoder neural networks on 4 datasets including simulated nonlinear data, dementia prognosis and diagnosis, and live-cell image analysis. Compared to prior techniques, ARMED models better distinguish confounded from true associations in simulations and learn more biologically plausible features in clinical applications. They can also quantify inter-cluster variance and visualize cluster effects in data. Finally, ARMED matches or improves performance on data from clusters seen during training (5-28% relative improvement) and generalization to unseen clusters (2-9% relative improvement) versus conventional models.
AB - Natural science datasets frequently violate assumptions of independence. Samples may be clustered (e.g., by study site, subject, or experimental batch), leading to spurious associations, poor model fitting, and confounded analyses. While largely unaddressed in deep learning, this problem has been handled in the statistics community through mixed effects models, which separate cluster-invariant fixed effects from cluster-specific random effects. We propose a general-purpose framework for Adversarially-Regularized Mixed Effects Deep learning (ARMED) models through non-intrusive additions to existing neural networks: 1) an adversarial classifier constraining the original model to learn only cluster-invariant features, 2) a random effects subnetwork capturing cluster-specific features, and 3) an approach to apply random effects to clusters unseen during training. We apply ARMED to dense, convolutional, and autoencoder neural networks on 4 datasets including simulated nonlinear data, dementia prognosis and diagnosis, and live-cell image analysis. Compared to prior techniques, ARMED models better distinguish confounded from true associations in simulations and learn more biologically plausible features in clinical applications. They can also quantify inter-cluster variance and visualize cluster effects in data. Finally, ARMED matches or improves performance on data from clusters seen during training (5-28% relative improvement) and generalization to unseen clusters (2-9% relative improvement) versus conventional models.
KW - Generalization
KW - biomedical imaging
KW - clinical data
KW - interpretability
KW - mixed effects model
KW - multilevel model
UR - http://www.scopus.com/inward/record.url?scp=85147309595&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85147309595&partnerID=8YFLogxK
U2 - 10.1109/TPAMI.2023.3234291
DO - 10.1109/TPAMI.2023.3234291
M3 - Article
C2 - 37018678
AN - SCOPUS:85147309595
SN - 0162-8828
VL - 45
SP - 8081
EP - 8093
JO - IEEE Transactions on Pattern Analysis and Machine Intelligence
JF - IEEE Transactions on Pattern Analysis and Machine Intelligence
IS - 7
ER -