[SciPy-user] Some mathematics/statisctics books
Wed Jun 11 04:20:17 CDT 2008
On Wed, Jun 11, 2008 at 8:03 AM, David Cournapeau
> didier rano wrote:
>> Wahh so many books to read.
>> But to analysis graph related to time-series, I don't know if I need
>> more a statistic approach or pure mathematic approach. Maybe that I
>> could use both approaches.
> It really depends on what you mean by pure mathematic approach. Purely
> mathematic approach to probabilities and statistics are mostly just
> that: purely mathematical. Don't get me wrong, maths is great, and
> probabilities/statistics are interesting mathematics topics on their
> own, but if you want to handle graphs, time series and all that, I don't
> think it will help you much.
> I second the book by Wasserman, although it does not treat a lot of time
> series stuff. But it is concise and precise (it is written with a
> relatively practical POV by someone who is definitely familiar with the
> theory; in particular, there are a lot of subtle examples and counter
> examples which are well explained, contrary to many other books).
> Another book. which I have not read entirely yet, but looks related to
> what you are looking for, is the book by Gelman et al.:
> "Bayesian Data Analysis", by Gelman A., John B. Carlin
> <http://www.rch.org.au/cebu/staff.cfm?doc_id=5690>, Hal S. Stern
> <http://www.ics.uci.edu/%7Esternh/>, and Donald B. Rubin.
> Not much theory there, but is really oriented toward data analysis as
> the title suggests :)
>> Thanks for all your help, and sorry for my poor background in
>> mathematics (I need to learn linear algebre too !)
> If you do multivariate analysis, you need to be more than familiar with
> linear algebra, I think. I don't know any good reference on this, but
> some open courseware may be nice (they have some video, too):
> SciPy-user mailing list
Fred Allen - "Imitation is the sincerest form of television."
More information about the SciPy-user