Webinar | RoboClinic Automating Machine Learning for Healthcare скачать в хорошем качестве

Webinar | RoboClinic Automating Machine Learning for Healthcare

For further information about this event or the upcoming KIKX events, visit our website: https://kikx.kfupm.edu.sa/pages/home/... Or follow our social media accounts:   / kfupm_kikx     / kfupm-kikx   This event was organized By SDAIA – KFUPM Joint Research Center for AI  Abstract: Autonomous Machine Learning (AutoML) involves the automation of core ML modules in a deployment and production environment. This talk presents RoboClinic, an innovative AutoML tool for clinical predictions, forecasts, and dashboards. Compared to ClosedLoop (the only existing Auto-ML tool for healthcare), RoboClinic also includes Auto-Business Intelligence (i.e., automatic creation of dashboards), creation of "hot" and "cold" ML sub-profiles during the automation, and ensemble methods of feature selection and algorithm selection. These features create much potential for RoboClinic's application to the healthcare community. RoboClinic has been tested on two local hospitals regarding Length of Stay predictions and mortality predictions of in-patients. The talk will detail RoboClinic's architecture and will discuss the current results, along with the process of its application to clinical data streams.  Speaker: Dr. Tariq Mahmood Dr. Tariq Mahmood, Professor and Coordinator of MS Programs, holds a Ph.D. Degree in Computer Science from the University of Trento, Italy. His career spans 13 years of experience in higher education research, teaching, and management. He has published over 60 research publications and has a citation score of ~1300 in Google Scholar. He has more than 8 years of experience in conducting training and consultancies in data analytics and related domains. He heads the Big Data Analytics laboratory at IBA with a focus on creating and maintaining academic-industry collaborations including ongoing projects with leading organizations such as Tabba Heart Institute (https://tabbaheart.org/), AKUH (https://hospitals.aku.edu/Pages/defau..., and 10pearls (https://10pearls.com/).