Interface Contraction in Biostatistics

Area: Nutritional Epidemiology
A specific automatic differentiation technique known as interface contraction is considered in the development of derivative codes for two loglikelihood functions used for biostatistical analysis. These functions were motivated by a problem in nutritional epidemiology investigating the relationships of age and dietary intakes of saturated fat, total energy, and alcohol with the fouryear risk of developing breast cancer in a prospective cohort of 89,538 nurses. The likelihood functions take into account measurement error in total energy and binary misclassication in saturated fat and alcohol intake. Newton's method with a line search globalization is employed for the optimization of these nonlinear likelihood functions, and hence the Hessian and gradient are needed at every optimization step. The Hessian is also needed to evaluate the variancecovariance matrix of the resulting maximum likelihood estimates.
Related ADTool:
ADIFOR
Reference:
People involved:
Reference:
Paul D. Hovland, Christian H. Bischof, Donna Spiegelman, Mario Casella
Efficient Derivative Codes through Automatic Differentiation and Interface Contraction: An Application in Biostatistics
Article in SIAM Journal on Scientific Computing, 1997
Efficient Derivative Codes through Automatic Differentiation and Interface Contraction: An Application in Biostatistics
Article in SIAM Journal on Scientific Computing, 1997
People involved:
 P. Hovland, C. Bischof,
Mathematics and Computer Science Division, Argonne National Laboratory  D. Spiegelman, M. Casella,
Harvard School of Public Health