Full Bayesian multi-task learning for multi-output brain decoding and accommodating missing data
Abstract
Multi-task learning (MTL) has recently been demonstrated to be highly promising for decoding multiple target variables from neuroimaging data. Its primary advantage is that it makes more efficient use of the data than existing decoding models, leading to improved accuracy. In this work, we propose a novel Bayesian MTL approach, motivated by problems such as clinical applications where accurate quantification of uncertainty is crucial. We present a Markov chain Monte Carlo approach to perform inference in the model and demonstrate the approach using a publicly available neuroimaging dataset. We study the conditions where MTL is likely to improve performance: we first evaluate MTL as an approach for accommodating missing data, which is an important problem that has received little attention from the neuroimaging community. We then examine whether it is beneficial to include classification and regression tasks in the same model. We relate our conclusions to results from geostatistics, where MTL methods were pioneered, and make recommendations for neuroimaging practitioners using MTL
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Gaussian process regression can be extended to directly predict multivariate phenotypes.
“Alternatively, Gaussian process regression can be extended to directly predict multivariate phenotypes, as we demonstrated previously (43, 44).”
From Understanding heterogeneity in clinical cohorts using normative models: Beyond case-control studies (2016) · cites this paper as [44]
Cites
Cressie, N. A. C. (1993) · Wiley Series in Probability and Statistics · doi:10.1002/9781119115151 · Cited by 6,507 · see citing works
Marcus, D. S., Wang, T. H., Parker, J., Csernansky, J. G., Morris, J. C., & Buckner, R. L. (2007) · Journal of Cognitive Neuroscience · doi:10.1162/jocn.2007.19.9.1498 · Cited by 2,163 · see citing works
Riemann manifold langevin and hamiltonian monte carlo methods
Girolami, M., & Calderhead, B. (2011) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.1467-9868.2010.00765.x · Cited by 1,473 · see citing works
Multi-task Gaussian Process Prediction
Bonilla, E. V., Chai, K. M. A., & Williams, C. K. I. (2007) · Cited by 916 · see citing works
Zhang, D., & Shen, D. G. (2011) · NeuroImage · doi:10.1016/j.neuroimage.2011.09.069 · Cited by 679 · see citing works
Kernels for Vector-Valued Functions: a Review
Álvarez, M. A., Rosasco, L., & Lawrence, N. D. (2012) · White Rose Research Online (University of Leeds, The University of Sheffield, University of York) · Cited by 587 · see citing works
Yuan, L., Wang, Y., Thompson, P. M., Narayan, V. A., & Ye, J. (2012) · NeuroImage · doi:10.1016/j.neuroimage.2012.03.059 · Cited by 195 · see citing works
Wang, H., Nie, F., Huang, H., Yan, J., Kim, S., Risacher, S. L., Saykin, A. J., & Shen, L. (2012) · Cited by 73 · see citing works
Bayesian multi-task learning for decoding multi-subject neuroimaging data
Marquand, A. F., Brammer, M., Williams, S. C. R., & Doyle, O. M. (2014) · NeuroImage · doi:10.1016/j.neuroimage.2014.02.008 · Cited by 59 · see citing works
Bayesian Multitask Classification With Gaussian Process Priors
Skolidis, G., & Sanguinetti, G. (2011) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.2011.2168568 · Cited by 57 · see citing works
Filippone, M., Marquand, A. F., Blain, C. R. V., Williams, S., Mourão-Miranda, J., & Girolami, M. (2012) · The Annals of Applied Statistics · doi:10.1214/12-aoas562 · Cited by 35 · see citing works
Self-citationFilippone, M., Marquand, A., Blain, C. R. V., Williams, S., Mourão-Miranda, J., Girolami, M., et al. (2011) · Europe PMC (PubMed Central) · Cited by 33 · see citing works
Self-citation
Flagged first, then by citations
Cited by
Sorted by citations
Understanding heterogeneity in clinical cohorts using normative models: Beyond case-control studies
Marquand, A. F., Rezek, I., Buitelaar, J., & Beckmann, C. F. (2016) · Biological Psychiatry · doi:10.1016/j.biopsych.2015.12.023 · Cited by 683 · see citing works
Self-citationA survey on multi‐output regression
Borchani, H., Varando, G., Bielza, C., & Larrañaga, P. (2015) · Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery · doi:10.1002/widm.1157 · Cited by 672 · see citing works
Predicting Alzheimer's disease progression using deep recurrent neural networks
Nguyen, M., He, T., An, L., Alexander, D. C., Feng, J., Yeo, B. T. T., & Initiative, F. T. A. D. N. (2020) · NeuroImage · doi:10.1016/j.neuroimage.2020.117203 · Cited by 216 · see citing works
Modeling longitudinal imaging biomarkers with parametric Bayesian multi‐task learning
Aksman, L., Scelsi, M. A., Marquand, A. F., Alexander, D. C., Sebastien Ourselin, A., Altmann, A., & for, A. D. N. I. (2019) · Human Brain Mapping · doi:10.1002/hbm.24682 · Cited by 18 · see citing works
Self-citation
- Authors
- Marquand, A. F., Williams, S. C. R., Doyle, O. M., & Rosa, M. J.
- Affiliations
- King's College London (GB)
- Venue
- 2014 International Workshop on Pattern Recognition in Neuroimaging
- Publisher
- IEEE
- Published
- 2014-06
- Volume
- 20
- Pages
- 1-4
- ISSN
- 2330-9989 (ISSN-L)
- OpenAlex
- W1998407126
- Cited by
- 4
- References
- 12
- Citations by year
- 2020: 1 · 2019: 1 · 2016: 1 · 2015: 1
- FWCI
- 1.24
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- International Workshop on Pattern Recognition in NeuroImaging (publishedVersion)
- Research Portal (King's College London) (submittedVersion)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › Functional Brain Connectivity Studies
- Physical Sciences › Computer Science › Artificial Intelligence › Gaussian Processes and Bayesian Inference
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Keywords
- multi-task learning, brain decoding, missing data, uncertainty quantification, Markov chain Monte Carlo, neuroimaging, classification and regression, geostatistics
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
