We will cover a variety of methods in temporal, spatial, and space-time statistics, learning to predict values at unobserved sites, model compositional data, looking at extreme values, decomposing data into different scales, etc. Among applications will be setting air quality standards, estimating air quality fields for health effect studies, studying the ecology of omnivores, looking at trends in climate data, and combining measurement from different instruments.

Event history analysis arises in many different contexts: health and medicine (survival analysis), engineering (reliability analysis), finance and actuarial science (risk analysis), and social sciences. This workshop will cover various models and methods for analyzing event history data. Topics will be selected from the following: hazard rates, shapes and interpretation; different types of censored data; parametric and nonparametric inference for time-to-failure distributions; proportional hazards and accelerated failure time regression models; point process models for time-between failures; and multi-state models. Applications and data sets from engineering, credit risk in finance, and survival analysis will be used extensively.

Students should have had a course at the undergraduate level in probability and statistical theory. Some familiarity with multivariate normal distribution, maximum likelihood estimation, regression analysis, and central limit theorem is desirable.