Data mining, statistics and R by Johannes Ledolter and Robert McCulloch

[rspa...@mcw.edu]

The Milwaukee Chapter of the ASA presents

The 2014 Annual Meeting

Data mining, statistics and R by Johannes Ledolter and Robert McCulloch

April 4, 2014 at UWM

Current members and students $40
New/former members           $49

You can make a reservation by sending a check made out to
the Milwaukee Chapter of the ASA to the Secretary/Treasurer:

Medical College of Wisconsin
Center for Patient Care & Outcomes Research
8701 Watertown Plank Rd.
Milwaukee, WI 53226

ATTN: Elizabeth Smith

Or via Paypal: els...@mcw.edu

You will receive an email confirmation and more details.
Please feel free to share this message with others who may be interested.

We also are holding a referendum; see the details at
http://community.amstat.org/MilwaukeeChapter/Home

Johannes Ledolter, PhD
C. Maxwell Stanley Professor of 
International Operations Management
Tippie College of Business
University of Iowa
http://www.biz.uiowa.edu/faculty/jledolter

Data mining and business analytics with big and small data
- it is not always size that matters 

Abstract: Will review useful methods for data mining 
and business analytics describing several applications 
and case studies where these methods have proven 
useful.  Also will discuss the importance of collecting 
one's own data through carefully designed statistical 
experiments.

References

Ledolter, J.: Data Mining and Business Analytics with R. 
Wiley & Sons, June 2013. See 
http://www.biz.uiowa.edu/faculty/jledolter/datamining/
for the R programs that are used in the book.

Ledolter, J. and Swersey, A.J.: Testing 1 - 2 - 3: 
Experimental Design with Applications in Marketing and 
Service Operations, Stanford University Press, 2007.


Robert E. McCulloch, PhD
Katherine Dusak Miller Professor of Econometrics and Statistics
Booth School of Business
University of Chicago
http://www.chicagobooth.edu/faculty/directory/m/robert-e-mcculloch
Data Mining with Bayesian Trees

Abstract: Tree regression and classification modeling play a
fundamental role in data mining. In addition to being an important
method in their own right, trees are often used in ensemble methods
such as bagging and boosting in which high dimensional models are
built up from many trees.  In this talk, we review practical Bayesian
approaches to data mining with trees and illustrate simple, easy to
use, software implementations which are available in R.

Fitting trees to data presents interesting challenges. There are many
possible tree structures so that searching the tree space for an
appropriate model is challenging. Bayesian inferential methods have
much to offer. The prior specification allows for parsimonious
regularization of the fit as well as nonparametric model
specification. Markov chain Monte Carlo provides a framework for
specifying stochastic searches in this vast model space. Recent
advances in parallel computing have made these Bayesian approaches
practical for large data sets.


08:30-09:00  Registration
09:00-09:15  Welcome
09:15-10:30  Robert McCulloch: Part I
10:30-11:00  Break
11:00-11:30  Robert McCulloch: Part II
11:30-12:00  Johannes Ledolter: Part I
12:00-01:00  Lunch on your own
01:00-02:30  Johannes Ledolter: Part II
02:30-03:00  Break
03:00-04:30  Johannes Ledolter: Part III
05:00-?      Dinner with the speakers: Restaurant TBD; please RSVP


More details can be found on our web page
http://community.amstat.org/MilwaukeeChapter/Home