{"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/autoprognosis-automated-clinical-prognostic","title":"AutoPrognosis: Automated Clinical Prognostic Modeling via Bayesian Optimization with Structured Kernel Learning","arxiv_id":"1802.07207","date":"2018-02-20","proceeding":"ICML 2018","authors":["Ahmed M. Alaa","Mihaela van der Schaar"],"abstract":"Clinical prognostic models derived from largescale healthcare data can inform\ncritical diagnostic and therapeutic decisions. To enable off-theshelf usage of\nmachine learning (ML) in prognostic research, we developed AUTOPROGNOSIS: a\nsystem for automating the design of predictive modeling pipelines tailored for\nclinical prognosis. AUTOPROGNOSIS optimizes ensembles of pipeline\nconfigurations efficiently using a novel batched Bayesian optimization (BO)\nalgorithm that learns a low-dimensional decomposition of the pipelines\nhigh-dimensional hyperparameter space in concurrence with the BO procedure.\nThis is achieved by modeling the pipelines performances as a black-box function\nwith a Gaussian process prior, and modeling the similarities between the\npipelines baseline algorithms via a sparse additive kernel with a Dirichlet\nprior. Meta-learning is used to warmstart BO with external data from similar\npatient cohorts by calibrating the priors using an algorithm that mimics the\nempirical Bayes method. The system automatically explains its predictions by\npresenting the clinicians with logical association rules that link patients\nfeatures to predicted risk strata. We demonstrate the utility of AUTOPROGNOSIS\nusing 10 major patient cohorts representing various aspects of cardiovascular\npatient care.","url_abs":"http://arxiv.org/abs/1802.07207v1","url_pdf":"http://arxiv.org/pdf/1802.07207v1.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":"autoprognosis-automated-clinical-prognostic","repo_url":"https://github.com/vanderschaarlab/autoprognosis","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"bayesian-optimization","task_name":"Bayesian Optimization"},{"task_slug":"diagnostic","task_name":"Diagnostic"},{"task_slug":"meta-learning","task_name":"Meta-Learning"},{"task_slug":"prognosis","task_name":"Prognosis"}],"methods":[{"method_slug":"gaussian-process","method_name":"Gaussian Process"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1802.07207","atlas_url":"https://app.syntology.ai/?focus=1802.07207","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}