Matthew J. Heaton
Ph.D. Student
Department of Statistical Science, Duke University


Environmental Protection Agency: September 2010-Present

In environmental science, computer models are frequently used to simulate environmental conditions. However, the computer simulations are often not well calibrated to conditions seen in the real world. I was hired by the EPA to write computer code to synthesize numerical model output and field measurments of various pollutants. In this way, the EPA can calibrate their computer models to closely resemble enivronmental conditions.

Denver Health and Hospital Authority: August 2008 - January 2011

Denver Health and Hospital Authority (DHHA) collects quarterly data from drug abuse programs across the United States. I was hired to develop a multivariate spatial-temporal conditional auto-regression (CAR) model for the purpose of tracking drug abuse cases over time. Additionally, using the statistical model, DHHA can provide accurate drug abuse reports to pharmaceutical companies.

Cormetech Inc: June 2009 - September 2009

Cormetech is a leading producer of selective catalytic reduction (SCR) systems for air pollution control. As part of the manufacturing process, researchers measure chemical concentrations in the system periodically. I was hired to develop an algorithm which alerts the researchers as to when the chemical concentrations have reached a steady state.

Statistical Consulting Center at Duke University: August 2008 - December 2009

I was a team member in the statistical consulting center within the Department of Statistical Science at Duke University. Specifically, students and professors from various departments within the Duke community would seek our aid on academic research projects.