ABSTRACT: Machine learning for health must be reproducible to ensure reliable clinical use. We evaluated 511 scientific papers across several machine learning subfields and found that machine learning for health compared poorly to other areas regarding reproducibility metrics, such as dataset and code accessibility. We propose recommendations to address this problem.
Read the full publication here.
ABSTRACT: Machine learning for health must be reproducible to ensure reliable clinical use. We evaluated 511 scientific papers across several machine learning subfields and found that machine learning for health compared poorly to other areas regarding reproducibility metrics, such as dataset and code accessibility. We propose recommendations to address this problem.
Read the full publication here.
ABSTRACT: Machine learning for health must be reproducible to ensure reliable clinical use. We evaluated 511 scientific papers across several machine learning subfields and found that machine learning for health compared poorly to other areas regarding reproducibility metrics, such as dataset and code accessibility. We propose recommendations to address this problem.
Read the full publication here.
Evidation and the Duke Big Ideas Lab have announced a partnership that aims to find ways to both increase representation of underserved populations and improve adherence among all participants in digital health studies.
The organizations are working together to build an analytic structure that will be available to other digital health researchers, helping to better predict study participation, adherence, and retention. Click below to read more about the project, led by Iredia M. Olaye, PhD, MSc, MHA, of Covered By Group and Evidation, in HealthLeaders.
Thanks to everyone who took the time to meet with us in San Francisco at this year’s J.P. Morgan Annual Conference and its surrounding events. Our team looks forward to staying in touch and helping you produce better health outcomes in 2023!
Bertina Yen is Evidation's Senior Vice President of People & Community, where she is responsible for the end-to-end employee experience and building a community that supports everyone's ability to show up as their authentic selves and perform their best work. Prior to joining Evidation, Bertina served as EVP overseeing clinical teams, product management, marketing, and product development at Zynx Health, driving development of evidence-based clinical decision support solutions that improve health and financial outcomes.
Bertina received her BAS from Stanford, then completed her MD and residency in internal medicine at UC San Diego. She also holds an MPH in health services administration from UCLA.
Bertina is the resident DJ at Evidation and sources the music that is played during the All-Employee Meetings. Outside of work, she enjoys hiking, traveling, and watching anything with a live audience (sporting events, plays, movies). She also starts every morning by completing at least one crossword puzzle!
Connect with Bertina here
ABSTRACT: Machine learning for health must be reproducible to ensure reliable clinical use. We evaluated 511 scientific papers across several machine learning subfields and found that machine learning for health compared poorly to other areas regarding reproducibility metrics, such as dataset and code accessibility. We propose recommendations to address this problem.
Read the full publication here.
ABSTRACT: Machine learning for health must be reproducible to ensure reliable clinical use. We evaluated 511 scientific papers across several machine learning subfields and found that machine learning for health compared poorly to other areas regarding reproducibility metrics, such as dataset and code accessibility. We propose recommendations to address this problem.
Read the full publication here.
ABSTRACT: Machine learning for health must be reproducible to ensure reliable clinical use. We evaluated 511 scientific papers across several machine learning subfields and found that machine learning for health compared poorly to other areas regarding reproducibility metrics, such as dataset and code accessibility. We propose recommendations to address this problem.
Read the full publication here.
McDermott M.B.A, Wang S, Marinsek N, Ranganath R, Foschini L, and Ghassemi M
ABSTRACT: Machine learning for health must be reproducible to ensure reliable clinical use. We evaluated 511 scientific papers across several machine learning subfields and found that machine learning for health compared poorly to other areas regarding reproducibility metrics, such as dataset and code accessibility. We propose recommendations to address this problem.
Read the full publication here.
