Statistics Triple Crown 2017

You are invited to participate in an afternoon of statistics seminars at NCAR.

Date: Thursday, July 13, 2017

Time: 1:00 pm - 4:00 pm

Location: National Center for Atmospheric Research in Boulder, Colorado

Mesa Lab, Main Seminar Room


PROGRAM (download a copy. pdf)

Each presentation will be followed by a short discussion and coffee break.

Mixture of Regression Models for Large Spatial Data Sets

Amanda Hering, Department of Statistical Science, Baylor University

When a spatial regression model that links a response variable to a set of
explanatory variables is desired, it is unlikely that the same regression
model holds throughout the domain when the spatial dataset is very large
and complex. The locations where the trend changes may not be known,
and we present here a mixture of regression models appraoch to 
identifying the locations wherein the relationships between the predictors
and the response is similar; to estimating the model within each group; and
to estimating the number of groups. An EM algorithm for estimating these
models is presented along with a criteria for choosing the number of groups.
An example with groundwater depth and associated predictors generated 
from a large physical model simulation demonstrates the fit and interpretation
of the proposed models.

Statistical Methods for Studying the West Antarctic Ice Sheet

Murali Haran, Department of Statistics , Penn State University

The melting of the West Antarctic ice sheet (WAIS) is likely to cause a 
significant rise in sea levels. Predicting the future behaviour of WAIS 
involves the use of computer models of ice sheet dynamics as well as 
ice sheet observational data. It is challenging to develop statistical methods
for such data because both the data and the computer modle output are in
the form of non-Gaussian spatial fields. I will describe an approach that 
combines Gaussian processes, generalized linear models, and dimension-
reduction approaches for spatial data. This approach allows for efficient
Markov chain Monte Carlo-based inference.  This is joint work with Won
Chang (U. of Cincinnati), Patrick Applegate and David Pollard (Penn State).

From Brain to Hand to Statistics with Dynamic Smoothing

Jim Ramsay, Michelle Carey, Juan Li
Department of Psychology, McGill University

Systems of differential equations are often used to model buffering 
processes that modulate a non-smooth high-energy input so as to produce
an output that is smooth and that distributes the energy load over time and
space. Handwriting is buffered in this way. We show that the smooth 
complex script that spells "statistics" in chinese can be represented as a
buffered version of a series of 46 equal-interval step inputs. The buffer 
consists of three undamped oscillating springs, one for each orthogonal
coordinate. The periods of oscillation vary slightly over the three coordinate
in a way that reflects the masses that are moved by muscle activations. Our
analyses of data on juggling three balls and on lip motion during speech 
confirm that this model works for a wide variety of human motions.

We use the term "dynamic smoothing" for the estimation of a structured
input functional object along with the buffer characteristics.


 After the seminar sessions, please join the group for an informal discussion at Under the Sun at 4:30pm.

Dorit Hammerling and Doug Nychka, Organizers