Re: [Phys-L] Model Selection Statistics

[brian whatcott <betwys1@sbcglobal.net>]

I sampled this work, and pleased with the informal yet comfortably 
authoritative tone, I ordered the cheapest available copy - $35 incl. 
shipping International ed. black & white.
Thanks Francois
Brian W


On 4/12/2019 4:51 PM, Francois Primeau wrote:
> Dear All,
> I teach a graduate level course on data analysis for Earth System Science
> Ph.D. students and I like the following text book:
>
>     Data Analysis A Bayesian Tutorial by D.S. Sivia and J. Skilling
>
> https://www.amazon.com/Data-Analysis-Bayesian-Devinderjit-Sivia/dp/0198568320
>
> Some of our students find the text book difficult, but that is because we
> have students with very varied backgrounds, from microbiology to
> mathematics and geography to geology and almost everything in between!  The
> students with undergraduate degrees in physics find the text book
> accessible. I like it very much.  Chapter 4 deals with model selection. It
> doesn't provide any recipes the way some statistics text books but it does
> discuss the issues in a logical way that is accessible to upper level
> undergraduate physics students.
>
> I think David MacKay's text book is also good
>      http://www.inference.org.uk/itprnn/book.html
> Chapter 28 discusses Model Comparison and Occam's Razor.
> Again no general recipe, but a clear discussion of the issues together with
> some useful mathematical approximations to simplify the analysis for some
> not too complicated models.
>
> Best,
> Francois
>
>
> On Fri, Apr 12, 2019 at 11:38 AM brian whatcott <betwys1@sbcglobal.net>
> wrote:
>
> > Starting with the "Math Is Fun" level:
> > https://www.mathsisfun.com/data/chi-square-test.html
> >
> > Continuing with the choices offered in the Indian journal of Ophthalmology:
> > https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3116565/
> >
> > Another Indian source for considering suitable tests:
> > https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2996580/
> >
> > Finally an Italian examination of how to validate competing models
> > https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2996580/
> >
> > They all seem to derogate categorical tests like Chi squared for
> > graph-like data sources of 80 points.
> >
> > Brian W
> > (not sure why American sources do not make a better
> > showing in my search?)
> >
> > On 4/12/2019 10:58 AM, Paul Nord wrote:
> > > "Ask n statisticians, get C(n,k) opinions."
> > >      - me, just now
> > >
> > > Suppose I have some physics data binned into 80 bins.  And I have two
> > > models which propose to fit some characteristic of interest and a
> > > non-linear background.  One is a simple polynomial.  The other is a more
> > > complicated.  I can calculate a reduced chi square for each data fit.
> > >
> > > What's the best way to compare the two models?
> > > I have a dim memory of this lecture.  Does someone have a good
> > translation
> > > of said lecture into terms a physicist can understand?
> > >
> > > Paul
> > > _______________________________________________
> > > Forum for Physics Educators
> > > Phys-l@mail.phys-l.org
> > > http://www.phys-l.org/mailman/listinfo/phys-l
> > _______________________________________________
> > Forum for Physics Educators
> > Phys-l@mail.phys-l.org
> > http://www.phys-l.org/mailman/listinfo/phys-l