Mathematical Statistics for High Dimension Low Sample Size Data J. S. Marron University of North Carolina A number of modern statistical applications involve data vectors where the dimension (number of variables), is much higher than the sample size (number of cases). This has motivated the development of statistical methods to handle such data (often not by statisticians). These advances have revealed a large new area in mathematical statistics. Perhaps surprisingly, asymptotics as the dimension goes to infinity, with the sample size fixed, turns out to produce useful insights. While the results and lessons learned go in new directions, relative to conventional asymptotics, this new way of thinking reveals insights into statistical inference that parallel the familiar approaches. Because this area is so new and unexplored, it is easy to make major contributions, with very little investment (although a willingness to think in non-standard ways is required). Attached are a couple of relevant papers. Best, Steve Naisyin Wang wrote: Dear all, thanks to our computing director, Dr. Schmiediche's help, now when you access the sub-directories through the link below the beautiful mountain picture of Colorado, you will be able to see the files therein besides being able to ftp materials in and out of there. For example, you can see Professor Simon Sheather's talk title & abstract in the "Title and Abstract" link (Title subdirectory). I hope for each research topics, if you have some materials to share, you can upload files there with WinSCP, http://winscp.net/eng/index.php userid: colorado passwd: David!Ruppert Please also put title and abstract of your presentation at the subdirectory "Title" or email it to me directly. This way I can ask for my student worker's help to get the files organized so that you can access from the main page. If there are errors or mistakes in the webpage, please kindly let me know so we can correct it. If you have not made the reservation, please kindly do so. I will double check with a few of you individually. When you made the reservation, please use the group code: CDPCTXM. I very much look forward to seeing you in less than a month. Naisyin On Tue, May 6, 2008 at 9:03 PM, Naisyin Wang > wrote: Dear all, I apologize for the delayed information. As you can expect, the student worker who has been helping me organize information was taking exams. We now have a webpage: http://www.stat.tamu.edu/~colorado/ The website is still under construction and we will keep updating it. * you can find driving direction there. * I will need your help to choose the boxlunch: please download the file http://www.stat.tamu.edu/~colorado/Information.doc fill it and send it back to me and Bhavika Shah >. * You should be able to WinSCP to the website: http://winscp.net/eng/index.php userid: colorado passwd: David!Ruppert You can --- put the title and abstract of your talk to html/Title --- put Research materials for "Measurement Errors Models" that you like to share to html/ME please also see incomplete info at http://www.stat.tamu.edu/~colorado/ME/infoME.html --- put Research materials for "Spatial Data" that you like to share to html/SP please also see incomplete info at http://www.stat.tamu.edu/~colorado/SP/infoSP.html --- put Research materials for "High Dimensional Data" that you like to share to html/HD please also see incomplete info at http://www.stat.tamu.edu/~colorado/HD/infoHD.html You can get to these webpages from the top of the main page (right below the picture of Keystone Mountains). I am learning from our computing director how to allow all to see the files in the directories. Please do not share the webpage with others -- it is for our convenient interactions only. We are also getting the ride information from all; once we have that, the info will be posted on the website. I very much look forward to seeing you. Naisyin [ Part 2, Application/PDF 349KB. ] [ Unable to print this part. ] [ Part 3, Application/PDF 215KB. ] [ Unable to print this part. ]