Dynamic Group Scheduling: Theory and Applications

在当今的商业世界中，管理人员生成和跟踪的数据比以往任何时候都多。这种信息收集技术改变了决策的方式;它与直觉无关，而更多地与数据分析有关。定量方法，帮助这一运动已成为关键工具，在实践中的循证决策。例如，2005年至2010年间，美国制造业的数据驱动决策增加了两倍。尽管如此丰富的信息已经揭示了奇妙的见解，但我们对相互依赖的系统做出决策的能力仍处于起步阶段。此外，使用大量数据来提出驱动这些系统的管理建议的定量方法非常缺乏。 我的研究计划的目的是研究新的定量工具来管理动态系统的不确定性。大部分现有文献都将这些系统分解为独立的、不相互作用的组件。每个组件负责处理一组任务;分析是在每个组件上独立执行的。然后在原始的未分解系统中实现所产生的见解，其中组件之间的交互通常被忽略或以特定的方式组合。然而，在实践中，适当地分离组件可能具有挑战性。系统的行为可能会有所不同，这取决于哪些任务是一起执行的，以及任务的处理顺序。 我的研究重点是开发定量方法来管理作为一个内聚单元的相互依赖的系统。它解决了物流、医疗保健和制造业的问题。在接下来的五年里，我将利用我在随机建模和数学优化方面的知识，以及我在分析复杂的现实世界过程方面的经验，为这些系统的管理员创建新的数学和计算工具。这项研究的资金将用于培训四名高素质的人员，将使我能够在北美的几个会议上介绍我的研究，并将相应的手稿提交给高影响力的同行评审期刊。我所提出的研究将为工业界和学术界做出有意义的贡献。此外，它具有很大的潜力，可以帮助加拿大成为开发数据驱动决策的定量方法的领导者之一。

In today's business world, managers generate and track more data than ever. This information gathering technology has changed the way decisions are made; it is less about intuition and more about data analysis. Quantitative methods that aid this movement have become key tools in the practice of evidence-based decision making. For example, data-driven decisions in US manufacturing tripled between 2005 and 2010. Although this abundance of information has uncovered fantastic insights, our ability to make decisions regarding inter-dependent systems is still in its infancy. Moreover, quantitative methods that use large amounts of data to make managerial recommendations that drive these systems are sorely lacking. The objective of my research proposal is to investigate new quantitative tools to manage dynamic systems with uncertainty. Much of the existing literature decomposes these systems into separate, non-interacting components. Each component is responsible for processing a set of tasks; analysis is performed on each component in isolation. The resulting insights are then implemented in the original non-decomposed system, where the interactions between components are usually ignored or combined in an ad-hoc fashion. In practice, however, it may be challenging to appropriately separate components. The system may behave differently depending on which tasks are performed together and the sequence the tasks are processed in. My research focuses on developing quantitative methods to manage inter-dependent systems as one cohesive unit. It tackles problems in logistics, healthcare, and manufacturing. Over the next five years, I will build on my knowledge of stochastic modeling and mathematical optimization and my experience analyzing complex, real-world processes, to create new mathematical and computational tools for administrators of these systems. Funding for this research will be used to train four high-quality personnel, will allow me to present my research at several conferences in North America, and submit the corresponding manuscripts to high-impact, peer-reviewed journals in operations research. My proposed research will make a meaningful contribution to both industry and the academic community. Furthermore, it has great potential to help establish Canada as one of the leaders in developing quantitative methodologies for data-driven decision-making.