Monte Carlo Statistical Methods

Yesterday, I started my annual course on Monte Carlo Statistical Method, open to the last year Master students in Paris Dauphine. This is a seven week course covering the first half and then more of the book Monte Carlo Statistical Method, with George Casella. Here are the (yes, 584) slides

These slides have not changed from last year (no time and no need to!) but this year I have distributed a preliminary copy of Introducing Monte Carlo Methods with R to the students so that they can practice Monte Carlo methods, spot typos in the book (only 10 cents per typo because I have no prior about the number of typos!), and write the(ir) solutions to the exercises as homework.

This entry was posted on November 10, 2009 at 12:22 am and is filed under Books, Statistics, University life with tags graduate course, Introducing Monte Carlo Methods with R, Monte Carlo Statistical Methods, simulation, slides. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

7 Responses to “Monte Carlo Statistical Methods”

Andrew’s criticisms « Xi'an's Og Says:
January 23, 2010 at 10:01 am

[…] a reader, you have to stop and grab it. Lots of fun.”) I noticed the same reactions from my students, so I quite agree with this point. When learning with a book, you need to sit with a piece of paper […]

Reply
jb Says:
December 10, 2009 at 3:28 pm

Dear Professor,

On page 39 of Monte Carlo Statistical Methods, for $u=0$ , the set of $x$ is $]-\infty,+\infty[$ . (1 can also be a trouble in some cases)
So either you extend the domain of $F$ to $-\infty$ and $+\infty$ or $u \neq 0,1$ as on wikipedia

My remark is not very interesting but in definition 2.3 the distribution function is defined on $\mathbb{R}$ . So it does not look very clean.

Best Regards

Reply
- xi'an Says:
  December 10, 2009 at 5:21 pm
  
  It is true that mathematically $u=0,1$ is an issue. But, probabilistically, generating a $\mathcal{U}(0,1)$ is equivalent to generating a $\mathcal{U}[0,1]$ in that 0 and 1 have a zero probability to occur. From a software point of view, good random generators cannot return those extreme values….
  
  Reply
xi'an Says:
November 20, 2009 at 7:37 am

The same student in my class pointed out further typos in Chapters 3 and 4. Fortunately, all those were signaled early enough to be corrected in the first printing.

Reply
xi'an Says:
November 10, 2009 at 4:59 pm

I already got two typos signaled by a student in the MCMC course:

– x[i,j]=x[i+n*(j-1)] instead of x[i+p*(j-1)] on page 10

– $F^{-}(u)=\inf\{x;F(x)\ge u\}$ instead of $F^{-}(u)=\inf\{x;F(x)\le u\}$ on page 44

Reply
alessandra Says:
November 10, 2009 at 11:27 am

Missed one seventh of the 5.84 hectoslides…!

Reply
Antoine Says:
November 10, 2009 at 1:59 am

Is it to make me feel guilty for not coming ?

Reply