Microsoft gives jailbreakers schwag
IronPython-based data/science environment
Unix as literature
Functional and imperative OOP
Breakthrough in partition theory: paper, lecture
Measure theory notes from Terry Tao
Power law distributions in empirical data
Four approaches to statistics
Interplay of Bayesian and Frequentist analysis
Labstats.net (Four approaches to Statistics) looks very useful, but they don’t have any contact information to report errors (It’s Perneger, nor Pernerger).
>Labstats.net (Four approaches to Statistics) looks very useful, but they don’t have any contact information to report errors.
Also, the Venn diagram says that there is no overlap between Bayesian statistics and information theory, which is wildly untrue.
Blaise: And in the “Interplay” link, Jim Berger argues that the Bayesian and Frequentist circles overlap (or should overlap) more than people realize. The quest to make a detailed diagram is doomed from the start, but it’s an interesting exercise.
It would also interesting to draw the circles so that their relative sizes carry some meaning. If you drew the circles in proportion to the fundamental importance of the concepts, the likelihood circle would be enormous. But if you drew the circles with sizes proportional to the number of professionals to identify with each, the likelihood circle would hardly be visible.
John D. Cook
Subscribe by Email