When can Multi-Site Datasets be Pooled for Regression? Hypothesis Tests, $\ell_2$-consistency and Neuroscience Applications

Hao Henry Zhou, Yilin Zhang, Vamsi K. Ithapu, Sterling C. Johnson, Grace Wahba, Vikas Singh
Proceedings of the 34th International Conference on Machine Learning, PMLR 70:4170-4179, 2017.

Abstract

Many studies in biomedical and health sciences involve small sample sizes due to logistic or financial constraints. Often, identifying weak (but scientifically interesting) associations between a set of predictors and a response necessitates pooling datasets from multiple diverse labs or groups. While there is a rich literature in statistical machine learning to address distributional shifts and inference in multi-site datasets, it is less clear when such pooling is guaranteed to help (and when it does not) – independent of the inference algorithms we use. In this paper, we present a hypothesis test to answer this question, both for classical and high dimensional linear regression. We precisely identify regimes where pooling datasets across multiple sites is sensible, and how such policy decisions can be made via simple checks executable on each site before any data transfer ever happens. With a focus on Alzheimer’s disease studies, we present empirical results showing that in regimes suggested by our analysis, pooling a local dataset with data from an international study improves power.

Cite this Paper

BibTeX
@InProceedings{pmlr-v70-zhou17c, title = {When can Multi-Site Datasets be Pooled for Regression? {H}ypothesis Tests, $\ell_2$-consistency and Neuroscience Applications}, author = {Hao Henry Zhou and Yilin Zhang and Vamsi K. Ithapu and Sterling C. Johnson and Grace Wahba and Vikas Singh}, booktitle = {Proceedings of the 34th International Conference on Machine Learning}, pages = {4170--4179}, year = {2017}, editor = {Precup, Doina and Teh, Yee Whye}, volume = {70}, series = {Proceedings of Machine Learning Research}, month = {06--11 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v70/zhou17c/zhou17c.pdf}, url = {https://proceedings.mlr.press/v70/zhou17c.html}, abstract = {Many studies in biomedical and health sciences involve small sample sizes due to logistic or financial constraints. Often, identifying weak (but scientifically interesting) associations between a set of predictors and a response necessitates pooling datasets from multiple diverse labs or groups. While there is a rich literature in statistical machine learning to address distributional shifts and inference in multi-site datasets, it is less clear when such pooling is guaranteed to help (and when it does not) – independent of the inference algorithms we use. In this paper, we present a hypothesis test to answer this question, both for classical and high dimensional linear regression. We precisely identify regimes where pooling datasets across multiple sites is sensible, and how such policy decisions can be made via simple checks executable on each site before any data transfer ever happens. With a focus on Alzheimer’s disease studies, we present empirical results showing that in regimes suggested by our analysis, pooling a local dataset with data from an international study improves power.} }
Endnote
%0 Conference Paper %T When can Multi-Site Datasets be Pooled for Regression? Hypothesis Tests, $\ell_2$-consistency and Neuroscience Applications %A Hao Henry Zhou %A Yilin Zhang %A Vamsi K. Ithapu %A Sterling C. Johnson %A Grace Wahba %A Vikas Singh %B Proceedings of the 34th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2017 %E Doina Precup %E Yee Whye Teh %F pmlr-v70-zhou17c %I PMLR %P 4170--4179 %U https://proceedings.mlr.press/v70/zhou17c.html %V 70 %X Many studies in biomedical and health sciences involve small sample sizes due to logistic or financial constraints. Often, identifying weak (but scientifically interesting) associations between a set of predictors and a response necessitates pooling datasets from multiple diverse labs or groups. While there is a rich literature in statistical machine learning to address distributional shifts and inference in multi-site datasets, it is less clear when such pooling is guaranteed to help (and when it does not) – independent of the inference algorithms we use. In this paper, we present a hypothesis test to answer this question, both for classical and high dimensional linear regression. We precisely identify regimes where pooling datasets across multiple sites is sensible, and how such policy decisions can be made via simple checks executable on each site before any data transfer ever happens. With a focus on Alzheimer’s disease studies, we present empirical results showing that in regimes suggested by our analysis, pooling a local dataset with data from an international study improves power.
APA
Zhou, H.H., Zhang, Y., Ithapu, V.K., Johnson, S.C., Wahba, G. & Singh, V.. (2017). When can Multi-Site Datasets be Pooled for Regression? Hypothesis Tests, $\ell_2$-consistency and Neuroscience Applications. Proceedings of the 34th International Conference on Machine Learning, in Proceedings of Machine Learning Research 70:4170-4179 Available from https://proceedings.mlr.press/v70/zhou17c.html.

Related Material