In this thesis, airline network revenue management problem is considered for the case with no cancellations and overbooking. In literature, there exist several approximate probabilistic and deterministic mathematical models developed in order to maximize expected revenue at the end of the reservation period. The aim of this study is to develop models considering also the risks involved in the proposed booking control policies. Two linear programming models are proposed which incorporate the variance of the revenue. The objective of the models is to effectively balance the tradeoff between the expectation and variance of the revenue. The performances of the proposed models are compared to the previous models through a numerical study. The seat allocations resulting from the mathematical models are used in a simulation model working with several booking control policies. The probability distributions of the revenues are investigated and the revenues are compared in terms of expectation, standard deviation, coefficient of variation and probability of poor performance. It is observed that the use of the proposed models decreases the variability of the revenue and thereby the risk of probability of poor performance. Also, the expected revenues obtained by implementing the solutions of the proposed models with nested booking control policies turn out to be higher than other probabilistic models as long as the degree of variance incorporation is within some interval. When compared with the deterministic models, the proposed models provides for the decision makers with alternative, preferable policies in terms of the expectation and the variability measures.

We consider the problem of staffing large-scale service centers with multiple customer classes and agent types operating under quality-of-service (QoS) constraints. We introduce formulations for a class of staffing problems, minimizing the cost of staffing while requiring that the long-run average QoS achieves a certain pre-specified level. The queueing models we use to define such service center staffing problems have random inter-arrival times and random service times. The models we study differ with respect to whether the arrival rates are deterministic or stochastic. In the deterministic version of the service center staffing problem, we assume that the customer arrival rates are known deterministically. It is computationally challenging to solve our service center staffing problem with deterministic arrival rates. Thus, we provide an approximation and prove that the solution of our approximation is asymptotically optimal in the sense that the gap between the optimal value of the exact model and the objective function value of the approximate solution shrinks to zero as the size of the system grows large. In our work, we also focus on doubly stochastic service center systems; that is, we focus on solving large-scale service center staffing problems when the arrival rates are uncertain in addition to the inherent randomness of the system's inter-arrival times and service times. This brings the modeling closer to reality. In solving the service center staffing problems with deterministic arrival rates, we provide a solution procedure for solving staffing problems for doubly stochastic service center systems. We consider a decision making scheme in which we must select staffing levels before observing the arrival rates. We assume that the decision maker has distributional information about the arrival rates at the time of decision making. In the presence of arrival-rate uncertainty, the decision maker's goal is to minimize the staffing cost, while ensuring the QoS achieves a given level. We show that as the system scales large in size, there is at most one key scenario under which the probability of waiting converges to a non-trivial value, i.e., a value strictly between 0 and 1. That is, the system is either over- or under-loaded in any other scenario as the size of the system grows to infinity. Exploiting this result, we propose a two-step solution procedure for the staffing problem with random arrival rates. In the first step, we use the desired QoS level to identify the key scenario corresponding to the optimal staffing level. After finding the key scenario, the random arrival-rate model reduces to a deterministic arrival-rate model. In the second step, we solve the resulting model, with deterministic arrival rate, by using the approximation model we point to above. The approximate optimal staffing level obtained in this procedure asymptotically converges to the true optimal staffing level for the random arrival-rate problem. The decision making scheme we sketch above, assumes that the distribution of the random arrival rates is known at the time of decision making. In reality this distribution must be estimated based on historical data and experience, and needs to be updated as new observations arrive. Another important issue that arises in service center management is that in the daily operation in service centers, the daily operational period is split into small decision time periods, for example, hourly periods, and then the staffing decisions need to be made for all such time periods. Thus, to achieve an overall optimal daily staffing policy, one must deal with the interaction among staffing decisions over adjacent time periods. In our work, we also build a model that handles the above two issues. We build a two-stage stochastic model with recourse that provides the staffing decisions over two adjacent decision time periods, i.e., two adjacent decision stages. The model minimizes the first stage staffing cost and the expected second stage staffing cost while satisfying a service quality constraint on the second stage operation. A Bayesian update is used to obtain the second-stage arrival-rate distribution based on the first-stage arrival-rate distribution and the arrival observations in the first stage. The second-stage distribution is used in the constraint on the second stage service quality. After reformulation, we show that our two-stage model can be expressed as a newsvendor model, albeit with a demand that is derived from the first stage decision. We provide an algorithm that can solve the two-stage staffing problem under the most commonly used QoS constraints. This work uses stochastic programming methods to solve problems arising in queueing networks. We hope that the ideas that we put forward in this dissertation lead to other attempts to deal with decision making under uncertainty for queueing systems that combine techniques from stochastic programming and analysis tools from queueing theory. / text

abstract: With the increased demand for genetically modified T-cells in treating hematological malignancies, the need for an optimized measurement policy within the current good manufacturing practices for better quality control has grown greatly. There are several steps involved in manufacturing gene therapy. These steps are for the autologous-type gene therapy, in chronological order, are harvesting T-cells from the patient, activation of the cells (thawing the cryogenically frozen cells after transport to manufacturing center), viral vector transduction, Chimeric Antigen Receptor (CAR) attachment during T-cell expansion, then infusion into patient. The need for improved measurement heuristics within the transduction and expansion portions of the manufacturing process has reached an all-time high because of the costly nature of manufacturing the product, the high cycle time (approximately 14-28 days from activation to infusion), and the risk for external contamination during manufacturing that negatively impacts patients post infusion (such as illness and death). The main objective of this work is to investigate and improve measurement policies on the basis of quality control in the transduction/expansion bio-manufacturing processes. More specifically, this study addresses the issue of measuring yield within the transduction/expansion phases of gene therapy. To do so, it was decided to model the process as a Markov Decision Process where the decisions being made are optimally chosen to create an overall optimal measurement policy; for a set of predefined parameters. / Dissertation/Thesis / Masters Thesis Industrial Engineering 2020

nÃo hà / Nosso principal objetivo neste estudo à investigar o papel da heterogeneidade na atualizaÃÃo, depois de um choque de informaÃÃo, do risco subjetivo sobre vitimizaÃÃo de homicÃdio. Nesse sentido, os dados utilizados neste trabalho tambÃm atestam a superestimaÃÃo do crime encontrada na literatura. A novidade à que os entrevistados receberam um choque de informaÃÃo que consiste na taxa oficial de homicÃdios, mas a grande maioria deles mantÃm a mesma percepÃÃo inicial. Ao propor um modelo de Update Bayesiano permitindo que nenhuma atualizaÃÃo fosse realizada, dois modelos foram desenvolvidos: um Tobit modificado e um modelo Hurdle de dois nÃveis. Assim como em estudos anteriores, nossos resultados mostraram que poderÃamos prosseguir com uma abordagem de Update Bayesiano. Ainda, quanto mais altas as respostas iniciais eram definidas, mais propensos os indivÃduos estavam em proceder uma mudanÃa de percepÃÃo. AlÃm disso, fundamentalmente, pudemos racionalizar a decisÃo de nÃo revisar as respostas seguindo um argumento de qualidade/credibilidade da informaÃÃo percebida. Descobrimos que os participantes mais velhos e as mulheres sÃo mais relutantes nÃo apenas em alterar as respostas iniciais, mas tambÃm na escolha do nÃvel da nova resposta, em caso de mudanÃa. Outra conclusÃo feita foi que o nÃvel educacional dos entrevistados era insignificante em nosso exercÃcio. De fato, o nÃvel educacional do entrevistador teve um papel fundamental em ambas decisÃes de mudanÃa e magnitude de revisÃo. Finalmente, nossos resultados tambÃm levantaram fortes evidÃncias sobre aspectos de homofilia. A ocorrÃncia de uma correspondÃncia em gÃnero entre entrevistadores e entrevistados teve o maior impacto sobre a decisÃo de mudar e na magnitude da atualizaÃÃo neste estudo. / Our main purpose in this study is to investigate the role of heterogeneity into the update of subjective homicide victimization risk after an informational shock. In this sense, the data used here also attests the crime overestimation found in the literature. The novelty is that our respondents faced an informational shock consisting in the official homicide rate, but the vast majority of them keeps the same initial perception. In proposing a Bayesian Update model allowing that no update takes place, two models were developed: a modified Tobit and a two-tiered Hurdle model. In accordance with previous papers, our results showed that we could proceed with a Bayesian Update approach. Also, the higher initial responses are set, more likely individuals are in proceeding a change in perceptions. Furthermore, fundamentally, we could rationalize a non-updating decision following a perceived informational quality/credibility argument. We found that older participants and females are more reluctant not only to change initial responses, but also to choose the level of the new response, in case of an update. In addition, respondentsâ level of education was insignificant in our exercise. In fact, interviewersâ level of education had a key role in both the changing and updating magnitude decisions. Finally, our results also raised strong evidence on homophily aspects. The occurance of a matching in gender between interviewers and interviewees had a major impact on the decision to change and in the magnitude of the update in this study.