Comments on Lindberg’s Correction for the Effects of Observer Variation on Probabilistic Diagnosis
Keyword(s):
Lindberg [2] proposes an adjustment of a probabilistic system for diagnosis in the light of known observer variation in the eliciting of symptoms and signs. He finds that his proposal leads to somewhat diffident predictions. We show that such an adjustment is only appropriate when the system is implemented with observers who are less reliable than those who created the data-base. In this case, with certain assumptions, an adjustment may be made that is less radical than that of Lindberg. A simple numerical solution for binary symptoms is provided.
1981 ◽
Vol 20
(03)
◽
pp. 163-168
◽
2008 ◽
Vol 1
(1)
◽
pp. 21-32
◽
1982 ◽
Vol 21
(03)
◽
pp. 137-142
◽
1998 ◽
Vol 8
(1)
◽
pp. 11-11
Keyword(s):
1993 ◽
Vol 4
(1)
◽
pp. 161-166
◽
2005 ◽
Vol 4
(1)
◽
pp. 129-129
1990 ◽
Vol 45
(5)
◽
pp. 676-676
◽
Keyword(s):