JSM 2012 Home

JSM 2012 Online Program

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

Online Program Home

Abstract Details

Activity Number: 401
Type: Contributed
Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
Sponsor: Section on Statistical Computing
Abstract - #304871
Title: The Spatial Lasso with Applications to Unmixing Hyperspectral Images
Author(s): Daniel Samarov*+ and Matthew Clarke and Ji Yoon Lee and David Allen and Maritoni Litorja and Jeeseong Hwang
Companies: NIST and National Institute of Standards and Technology and National Institute of Standards and Technology and National Institute of Standards and Technology and National Institute of Standards and Technology and National Institute of Standards and Technology
Address: 100 Bureau Drive, Gaithersburg, MD, 20899, United States
Keywords: Multivariate statistics ; Sparse regression ; spatial statistics ; lasso ; splasso ; hyperspectral imaging
Abstract:

An important aspect of any statistical analysis is ensuring model prediction accuracy and identifying the variables most important to accurate prediction. The process of variable selection is particularly important when the true, underlying model has a sparse representation, as the inclusion of irrelevant variables can lead to inaccurate inference. As data sets have grown in size and complexity the class of l1-based models have gained popularity, with one of the best known being the LASSO. In this talk we propose a new regularization and variable selection model called the spatial LASSO (SPLASSO). The SPLASSO incorporates spatial information to help improve the model estimation process for multivariate data. A natural application for the methods developed here is hyperspectral imaging (HSI), a spectroscopic imaging technique which collects densely sampled measurements along the electromagnetic spectrum. We show how the SPLASSO can be used to determine the type and quantity of chemicals in these images and its advantages over several existing methods.


The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.

Back to the full JSM 2012 program




2012 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.