{"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/deep-multi-output-forecasting-learning-to","title":"Deep Multi-Output Forecasting: Learning to Accurately Predict Blood Glucose Trajectories","arxiv_id":"1806.05357","date":"2018-06-14","proceeding":null,"authors":["Ian Fox","Lynn Ang","Mamta Jaiswal","Rodica Pop-Busui","Jenna Wiens"],"abstract":"In many forecasting applications, it is valuable to predict not only the\nvalue of a signal at a certain time point in the future, but also the values\nleading up to that point. This is especially true in clinical applications,\nwhere the future state of the patient can be less important than the patient's\noverall trajectory. This requires multi-step forecasting, a forecasting variant\nwhere one aims to predict multiple values in the future simultaneously.\nStandard methods to accomplish this can propagate error from prediction to\nprediction, reducing quality over the long term. In light of these challenges,\nwe propose multi-output deep architectures for multi-step forecasting in which\nwe explicitly model the distribution of future values of the signal over a\nprediction horizon. We apply these techniques to the challenging and clinically\nrelevant task of blood glucose forecasting. Through a series of experiments on\na real-world dataset consisting of 550K blood glucose measurements, we\ndemonstrate the effectiveness of our proposed approaches in capturing the\nunderlying signal dynamics. Compared to existing shallow and deep methods, we\nfind that our proposed approaches improve performance individually and capture\ncomplementary information, leading to a large improvement over the baseline\nwhen combined (4.87 vs. 5.31 absolute percentage error (APE)). Overall, the\nresults suggest the efficacy of our proposed approach in predicting blood\nglucose level and multi-step forecasting more generally.","url_abs":"http://arxiv.org/abs/1806.05357v1","url_pdf":"http://arxiv.org/pdf/1806.05357v1.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":"deep-multi-output-forecasting-learning-to","repo_url":"https://github.com/igfox/multi-output-glucose-forecasting","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}