We propose the “study strap ensemble,” which combines advantages of two common approaches to fitting prediction models when multiple training datasets (“studies”) are available: pooling studies and fitting one model vs. averaging predictions from multiple models each fit to individual studies. The study strap ensemble fits models to bootstrapped datasets or “pseudo-studies.” These are generated by resampling from multiple studies with a hierarchical resampling scheme that generalizes the randomized cluster bootstrap. The study strap is controlled by a tuning parameter that determines the proportion of observations to draw from each study. When the parameter is set to its lowest value, each pseudo-study is resampled from only a single study. When it is high, the study strap ignores the multistudy structure and generates pseudo-studies by merging the datasets and drawing observations like a standard bootstrap. We empirically show the optimal tuning value often lies in between and prove that special cases of the study strap draw the merged dataset and the set of original studies as pseudo-studies. We extend the study strap approach with an ensemble weighting scheme that utilizes information in the distribution of the covariates of the test dataset.
Our work is motivated by neuroscience experiments using real-time neurochemical sensing during awake behavior in humans. Current techniques to perform this kind of research require measurements from an electrode placed in the brain during awake neurosurgery and rely on prediction models to estimate neurotransmitter concentrations from the electrical measurements recorded by the electrode. These models are trained by combining multiple datasets that are collected in vitro under heterogeneous conditions in order to promote accuracy of the models when applied to data collected in the brain. A prevailing challenge is deciding how to combine studies or ensemble models trained on different studies to enhance model generalizability.
Our methods produce marked improvements in simulations and in this application. All methods are available in the CRAN package.
GCL was supported by the NIH, F31DA052153; T32 AI 007358. GP and PP received support from NSF Grant DMS-1810829. KK received support from the NIH, R01 DA048096; R01 MH121099; R01 NS092701; 5KL2TR00142; WFSOM, Phys/Pharm & Neurosurgery.
The authors would like to thank the reviewers, the Associate Editor, and the Editor for their feedback that substantially improved the quality of this paper.
Gabriel Loewinger. Prasad Patil. Kenneth T. Kishida. Giovanni Parmigiani. "Hierarchical resampling for bagging in multistudy prediction with applications to human neurochemical sensing." Ann. Appl. Stat. 16 (4) 2145 - 2165, December 2022. https://doi.org/10.1214/21-AOAS1574