Thesis (M.Sc.). / Written for the School of Computer Science. Title from title page of PDF (viewed 2008/07/30). Includes bibliographical references.

Thesis (M.S.)--Auburn University, 2008. / Abstract. Vita. Includes bibliographical references (p. 36-37).

Thesis (Ph. D.)--University of Chicago, Dept. of Economics, June 1999. / Includes bibliographical references. Also available on the Internet.

Thesis (M.S. in Operations Research)--Naval Postgraduate School, Dec. 2004. / Thesis Advisor(s): James N. Eagle, W. Matthew Carlyle. Includes bibliographical references (p. 35). Also available online.

Thesis (Ph. D.)--Brigham Young University. Dept. of Electrical and Computer Engineering, 2006. / Includes bibliographical references (p. 111-118).

Thesis (Ph. D.)--Carnegie Mellon University, 2005. / "May 2005." Includes bibliographical references.

Thesis (M.S.)--Ohio State University, 1986. / Includes bibliographical references (leaves 89-90). Available online via OhioLINK's ETD Center

Thesis (M.S.)--University of Delaware, 2009. / Principal faculty advisor: Xiaoming Li, Dept. of Electrical & Computer Engineering. Includes bibliographical references.

This thesis studies two optimization problems: the optimization of a staffing policy assuming non stationary Poisson demand, and exponential travel and job times, and the optimization of investment decisions with an investment lag. In the staffing policy optimization, we solve a novel time-dynamic Hamilton-Jacobi-Bellman equation that models jobs as a Poisson jump process. The model gives the employer the flexibility to control the number of staff hired by two factors: the cost of hiring and the effect of delay. We have solved the optimal staffing policy problem using different approaches, which are compared. We produce accurate numerical results for different parameters, and discuss the advantages and disadvantages of each approach. Moreover, we have solved a staffing problem for a national utility company, using a standard linear programming approach, which is compared with our methods. In addition to the Poisson jump process, we extend the model to treat a continuous job model, and two locations model that is extendible to a larger network problem. In the investment lag problem, we use a mixture of numerical methods including finite difference and body fitted co-ordinates to form a robust and stable numerical scheme which is applied to solve the investment lag problem for a geometric Brownian motion presented in the paper by Bar-Ilan and Strange (1996). The problem is to calculate the optimal price to invest in a project that have a time lag period between the decision to invest and production, and the optimal price to mothball the project. The method presented in this thesis is more flexible as we compare it with the previous results, and solves the problem for different stochastic processes, such as Cox-Ingersoll-Ross model, which does not have analytic solution.

Dynamic programming (DP) has long been an essential framework for solving sequential decision-making problems. However, when the state space is intractably large or the objective contains a risk term, the conventional DP framework often fails to work. In this dissertation, we investigate such issues, particularly those arising in the context of multi-armed bandit problems and risk-sensitive optimal execution problems, and discuss the use of modern DP techniques to overcome these challenges such as information relaxation, policy gradient, and state augmentation. We develop frameworks formalize and improve existing heuristic algorithms (e.g., Thompson sampling, aggressive-in-the-money trading), while shedding new light on the adopted DP techniques.