We are developing novel methods to accurately predict cancer patient outcomes by applying multitask learning algorithms to information shared across multiple cancer types to jointly learn commonalities and improve model predictive performance. Not only will this technology improve precision oncology, but it is transferable to other fields of research.
Goncalves, A., et al. 2019. "Bayesian Multitask Learning Regression for Heterogeneous Patient Cohorts." Journal of Biomedical Informatics 4:1–10. LLNL-JRNL-767154.
Lawrence Livermore National Laboratory • 7000 East Avenue • Livermore, CA 94550
Operated by Lawrence Livermore National Security, LLC, for the Department of Energy's National Nuclear Security Administration.