This work is motivated by the appointment booking of specialist services at the ambulatory care center (ACC) in a major public hospital in Hong Kong. We study scheduling policies in a healthcare system in which patients' waiting times for medical consultations are governed by the Hong Kong Hospital Authority. Specifically, the ACC accepts only advanced bookings, and patients should receive consultation within a stipulated target waiting time. We study two forms of scheduling: allocation scheduling and advance scheduling.

We consider the problem of allocating daily hospital service capacity among several types of elective surgical procedures. Our focus is on the interaction between two major constraining hospital resources: operating room and recovery bed capacity. In our model, each type of surgical procedure has an associated revenue, stochastic procedure duration and stochastic length-of-stay.

We study stochastic optimization problems with decisions truncated by random variables. The technical difficulty is that the optimization problems may not be convex programs due to the truncation. We develop a transformation technique to convert the original non-convex problems to equivalent convex ones. Our transformation allows us to prove the preservation of some desired structural properties, such as convexity, submodularity, and L-natural-convexity, under optimization operations, that are critical for identifying the structures of optimal policies and developing efficient algorithms.

Opponents of pollution regulation claim that regulations choke businesses, hurt welfare and destroy economic growth. Do pollution regulations in fact hurt producers and the economy? To address this question, we develop an integrated production-pollution-abatement model, and study two popular regulations-- Cap-and-Trade and Taxes-- under imperfect competition. We show that firms' strategic tradeoffs between output reduction and pollution abatement, not identified in the prior literature, drive industry profits, consumer surplus and welfare.

Using network survey and objective productivity data collected from the Esquel Group, a large textile company headquartered in Hong Kong, we find that the density of worker social connections within a production line positively and quadratically affects the line-average productivity. However, we find that at individual level social connections do not affect the mean productivity. Instead, the density of social connections affects the variance of productivity at individual level, which reasonably explains the quadratic pattern of influence at the line level.

This paper provides an examination of the notion of “distance” and ordering between the whole distributions of earnings for men and women, and some related counterfactuals.We investigate aggregative measures and their statistical implementation and evaluation. We provide a synthesis of the decision theoretic basis of measures for such things as the “gender gap”. “Strong” comparison/ordering of distributions of outcomes requires subjective evaluation functions reflecting decision maker’s preferences. Examples include popular notions of the “gender gap” based on means or medians.

Deep learning is a form of machine learning for nonlinear high dimensional pattern matching and prediction. By taking a Bayesian probabilistic perspective, we provide a number of advantages, with more efficient algorithms for optimization and hyper-parameter tuning, and an explanation of predictive performance. A framework for constructing good Bayesian predictors in high dimensions is provided.

This presentation gives an overview of recent developments on the Bayes risk in new contexts such as ranking forecast models by stochastic error distance (SED) proposed by Diebold and Shin (2017), Gini coefficient of income inequality, and expected Fisher information. The mean excess or residual function is the optimal predictor under the quadratic loss of the amount a random variable exceeds a given threshold, which is a function of the threshold, hence a local measure. Its global risk under a distribution for the threshold is the Bayes risk.

The agriculture industry plays a critical role in the U.S. economy and various industry sectors depend on the output of farms. To protect and raise farmers' income, the U.S. government offers two subsidy programs to farmers: the Price Loss Coverage (PLC) program which pays farmers a subsidy when the market price falls below a reference price, and the Agriculture Risk Coverage (ARC) program which is triggered when farmers' revenue is below a threshold. Given the unique features of PLC and ARC, we develop models to analyze their impacts on consumers, farmers, and the government.

In this paper we propose a novel misspecification test for simulation metamodels. It is a consistent test that helps to assess the adequacy of simulation metamodels. The test statistic we construct is shown to be asymptotically normally distributed under the null hypothesis that the metamodel is correct, while diverging to infinity at a rate of√n where n is the test sample size if the given metamodel is inadequate. Furthermore, as a by-product, we construct confidence intervals for mean squared errors of the metamodels.