31 - 40 of 44 results for: OIT

This course provides an overview of research in supply chain management (SCM). It has three parts. The first part reviews basic tools of SCM research through selected readings in economics, IT and operations research. The second part reviews the literature in SCM, covering topics such as inventory models, information sharing, information distortion, contract design, value of integration, performance measurement, risk management, and the use of markets for procurement. The last part is devoted to recent advances in SCM research.

Description is currently unavailable because of ongoing review of the OIT PhD program by OIT faculty. Description will become available when the review is completed at the end of the Summer.

This course covers basic analytical tools and methods that can be used in research in operations and information systems. The emphasis is on foundations of stochastic inventory theory. Basic topics include convexity, duality, induced preference theory, and structured probability distributions. Much of the course is devoted to Markov decision processes, covering finite and infinite horizon models, proving the optimality of simple policies, bounds and computations, and myopic policies.

Queueing models may be used to represent service delivery systems, manufacturing processes, or data processing networks. The first half of this two-unit course consists of lectures on performance analysis (e.g., estimating congestion and delay) and control of queueing systems using asymptotic methods, both in the traditional heavy traffic regime and in the Halfin-Whitt regime. The second half consists of student presentations of recent papers in asymptotic methods in queueing systems. Prerequisites: Statistics 217 and 218, or consent of instructor; some prior exposure to stochastic models in general, and queueing theory in particular, is useful but not essential.

This seminar will highlight the research evolution and advances on the smart use of information in supply chain management. Advances in technologies like real-time information systems, decision support methodologies, the internet and mobile technologies such as RFID (radio-frequency identification) have also enabled visibility and structural changes that result in significant supply chain performance enhancements. In parallel to the development of new practices and concepts in industry, we will examine emerging research that are based on (1) structuring new processes and supply chain networks with the new technologies; (2) exploring ways to do planning and make decisions consequently; (3) quantifying the benefits as a result; and (4) aligning the incentives of multiple players in a supply chain when the costs and benefits to these players are different.

The goal of this course is to provide a comprehensive introduction to the theory and practice of revenue management. It will comprise of a set of lectures that will cover the theoretical fundamentals of the area as well as an overview of current research developments through the presentation and discussion of recent papers. Topics include capacity control (single-resource and network), consumer behavior and market response models, dynamic pricing, procurement auctions, price experimentation, supply chain management and pricing.

The first half of this course will cover (i) the basic theory of Brownian motion, (ii) Ito stochastic calculus, and (iii) the rudiments of continuous-time stochastic control, all undertaken at a brisk pace, aimed at students who already know the basics or else have a strong enough math background to learn them quickly. The text for this part of the course will be Brownian Models of Performance and Control, by J. Michael Harrison, Cambridge University Press, 2013, which can be ordered from Amazon: http://www.amazon.com/Brownian-Performance-Control-Michael-Harrison/dp/1107018390/ref=sr_1_1?ie=UTF8&qid=1395420072&sr=8-1&keywords=Brownian+Models+of+Performance+and+ControlnnnThe second half of the course will explore in depth some models arising in operations research, finance and economic theory, such as the McDonald-Siegel investment model (an optimal stopping problem, treated in Chapter 5 of the textbook), Brownian versions of the classic cash balance problem (a family of stochastic control problems, treated in Chapter 7 of the textbook), and Yuliy Sannikov’s continuous-time principal-agent model (Review of Economic Studies, 2008). The course will be rather informally organized, more of a collaboration between students and instructor than a top-down lecture format, with at least half of the class time devoted to presentation of problems by students and auditors.

This course aims to introduce students to research topics in data-driven decision making with specific attention to healthcare applications. However, most concepts are applicable in areas beyond healthcare as well. Examples of topics are: prediction and risk adjustment, computational and statistical challenges associated with large-scale data, and dynamic decision making under uncertainty.

In most real-world problems, decision-makers often face uncertainty with respect to the underlying modelsnnthat drive the rewards/costs associated with potential strategies. The uncertainty in the problem can bennmodeled in a number of ways (e.g., a probability distribution over some parameters or an uncertainty setnnfor some variables) and a selection of an appropriate framework depends on various considerations rangingnnfrom the availability of historical data (or lack thereof) to the robustness of resulting strategies or thenntractability of the formulation. In addition, once a framework is selected, further challenges often arise whennnconsidering dynamic settings, in which the level of uncertainty may be updated from one period to another.nnThe high-level objectives of this course are:nn1. to introduce various frameworks for decision-making under model uncertaintynn2. to introduce tools to solve such problems, including ones to develop optimal or near-optimal learningnnstrategiesnn3. to discuss the various tradeoffs that arise such as tractability vs. performance, exploration vs. exploitation, and remembering vs. forgettingnn4. to explore research papers that demonstrate applications of discussed methods and models to variousnnproblems areas such as dynamic pricing, revenue management, inventory management, and assortmentnnselection