Semiparametric methods continue to be an active research area, especially as computing resources and power grow. These methods are used for inference in models with both parametric and nonparametric components. Work in this setting ranges from semiparametric modeling and efficiency to semiparametric regression. The former involves making efficient inferences in the presence of infinite dimensional nuisance parameters. The latter corresponds to regression models with parametric errors and unspecified (nonparametric) regression functions. Fundamental work remains to be done in high dimensional missing data problems, both in the setting of frequentist and Bayesian inference and in estimating regression functions with the complications posed in many different areas of application. The workshop brings together leaders in the field of semiparametric methodology and young researchers. The invited speakers all are very distinguished individuals who have done outstanding work in semiparametrics and have co-authored three of the seminal texts.
Flexible modeling with few assumptions is an integral component of solving many scientific problems. Semiparametric methods are a way to accomplish this. Major advances have been made in the past thirty years. However, fundamental work remains to be done in high dimensional problems with incomplete data, in estimating regression functions with the complications posed in many different areas of application, among other areas. Applications for such work are wide ranging and include spatio-temporal models, capture-recapture models, and measurement error models in environmental applications, classification and testing in in genetics and genomics, and missing data models in longitudinal observational studies and clinical trials. The workshop provides an excellent opportunity to discuss the many recent significant developments in semiparametric methodology and to identify important problems and new research directions.
