Vocabulary:
lcsh
Type:
Topic
Note:
Work cat.: 2004059195: Wallace, C.S. Statistical and inductive inference by minimum message length, 2005: CIP galley ("...
it became clear that message-length tools used in the classification method could be generalized to apply to many model-selection
and statistical inference problems, leading to our first attempts to formalize the minimum message length method. However,
these attempts seemed to be incomprehensible or repugnant to the referees of statistical journals. Fortunately, Peter Freeman,
a proper statistician who had looked at the stone circle problem, saw some virtue in the approach and very kindly spent a
year's sabbatical helping to frame the idea in acceptable statistical terms, leading to the first publication of MML in a
statistical journal.")
Note:
Wikipedia, Oct. 25, 2004 (Minimum message length; a formal information theory restatement of Occam's razor)
Note:
Web of science, Oct. 25, 2004.