BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:From bed to bench side: bringing machine learning to day-to-day cl
 inical practice - Dr Zohreh Shams (JRF\, Wolfson College\, University of C
 ambridge)
DTSTART:20201120T180000Z
DTEND:20201120T190000Z
UID:TALK153628@talks.cam.ac.uk
CONTACT:Julian Siebert
DESCRIPTION:Machine learning (ML) and in particular deep neural networks (
 DNNs) have the potential to transform diagnosis\, prognosis and treatment 
 planning in oncology. Uninterpretability of DNNs\, however makes their dep
 loyment in safety critical domains\, such as oncology\, challenging. This 
 is evidenced by the scepticism of clinical community about ML systems. In 
 this talk\, I present REM\, a model extraction approach that allows extrac
 ting rules from deep neural networks.\nUnlike interpretability methods\, s
 uch as feature importance and sample importance\, model extraction allows 
 human-simulatability that is an especially valuable feature for an interpr
 etability method suited for use in clinical domain. It allows clinicians t
 o inspect the predictions of a ML model and contrast them with their exper
 t knowledge to verify the biological relevance as well as to identify the 
 unintended bias. It further allows checking the impact of perturbation in 
 input on the output and adjusting the model accordingly.\n\nRegister "here
 ":https://wolfson-cam-ac-uk.zoom.us/webinar/register/1016038025751/WN_rgBK
 cjkkQX-m_fMfSZrUzQ
LOCATION:Online
END:VEVENT
END:VCALENDAR