{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/bayesian-inference-of-individualized","title":"Bayesian Inference of Individualized Treatment Effects using Multi-task Gaussian Processes","arxiv_id":"1704.02801","date":"2017-04-10","proceeding":"NeurIPS 2017 12","authors":["Ahmed M. Alaa","Mihaela van der Schaar"],"abstract":"Predicated on the increasing abundance of electronic health records, we\ninvesti- gate the problem of inferring individualized treatment effects using\nobservational data. Stemming from the potential outcomes model, we propose a\nnovel multi- task learning framework in which factual and counterfactual\noutcomes are mod- eled as the outputs of a function in a vector-valued\nreproducing kernel Hilbert space (vvRKHS). We develop a nonparametric Bayesian\nmethod for learning the treatment effects using a multi-task Gaussian process\n(GP) with a linear coregion- alization kernel as a prior over the vvRKHS. The\nBayesian approach allows us to compute individualized measures of confidence in\nour estimates via pointwise credible intervals, which are crucial for realizing\nthe full potential of precision medicine. The impact of selection bias is\nalleviated via a risk-based empirical Bayes method for adapting the multi-task\nGP prior, which jointly minimizes the empirical error in factual outcomes and\nthe uncertainty in (unobserved) counter- factual outcomes. We conduct\nexperiments on observational datasets for an inter- ventional social program\napplied to premature infants, and a left ventricular assist device applied to\ncardiac patients wait-listed for a heart transplant. In both experi- ments, we\nshow that our method significantly outperforms the state-of-the-art.","url_abs":"http://arxiv.org/abs/1704.02801v2","url_pdf":"http://arxiv.org/pdf/1704.02801v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"bayesian-inference-of-individualized","repo_url":"https://github.com/vanderschaarlab/mlforhealthlabpub/tree/main/alg/causal_multitask_gaussian_processes_ite","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"jax","reach":null}],"tasks":[{"task_slug":"bayesian-inference","task_name":"Bayesian Inference"},{"task_slug":"gaussian-processes","task_name":"Gaussian Processes"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"},{"task_slug":"selection-bias","task_name":"Selection bias"},{"task_slug":null,"task_name":"counterfactual"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1704.02801","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}