BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Algorithmic diagnosis and the problem of statistical evidence - Su
 ne Hannibal Holm (University of Copenhagen)
DTSTART:20240425T143000Z
DTEND:20240425T160000Z
UID:TALK215992@talks.cam.ac.uk
CONTACT:Lewis Bremner
DESCRIPTION:Disease detecting algorithms with superior accuracy are becomi
 ng available. From a societal point of view\, higher diagnostic accuracy c
 ould enable more efficient use of healthcare resources. From a patient poi
 nt of view\, higher accuracy could improve health outcomes for patients. I
 f we can reasonably expect that an algorithmic diagnostic procedure will b
 e more accurate than one involving human diagnosticians\, it is pertinent 
 to ask whether we should sometimes substitute algorithms for human diagnos
 ticians? In this talk I discuss whether there is a principled epistemologi
 cal argument against a diagnostic procedure that relies wholly on a diseas
 e detection algorithm shown to have superior accuracy. My discussion focus
 es on the purely statistical nature of algorithmic diagnosis and connects 
 algorithmic diagnosis to the extensive literature on the epistemology of s
 tatistical evidence and its use in court.
LOCATION:Large Lecture Theatre\, Department of Plant Sciences\, Downing Si
 te
END:VEVENT
END:VCALENDAR