Date of Award
Doctor of Philosophy in Computing Sciences - (Ph.D.)
Senjuti Basu Roy
Chase Qishi Wu
An emerging trend is to leverage human capabilities in the computational loop at different capacities, ranging from tapping knowledge from a richly heterogeneous pool of knowledge resident in the general population to soliciting expert opinions. These practices are, in general, termed human-in-the-loop (HITL) computations.
A HITL process requires holistic treatment and optimization from multiple standpoints considering all stakeholders: a. applications, b. platforms, c. humans. In application-centric optimization, the factors of interest usually are latency (how long it takes for a set of tasks to finish), cost (the monetary or computational expenses incurred in the process), and quality of the completed tasks. Platform-centric optimization studies throughput, or revenue maximization, while human-centric optimization deals with the characteristics of the human workers, referred to as human factors, such as their skill improvement and learning, to name a few. Finally, fairness and ethical consideration are also of utmost importance in these processes./p>
This dissertation aims to design solutions for each of the aforementioned stakeholders. The first contribution of this dissertation is the study of recommending deployment strategies for applications consistent with task requesters’ deployment parameters. From the worker’s standpoint, this dissertation focuses on investigating online group formation where members seek to increase their learning potential via collaboration. Finally, it studies how to consolidate preferences from different workers/applications in a fair manner, such that the final order is both consistent with individual preferences and complies with a group fairness criteria.
The technical contributions of this dissertation are to rigorously study these problems from theoretical standpoints, present principled algorithms with theoretical guarantees, and conduct extensive experimental analysis using large-scale real-world datasets to demonstrate their effectiveness and scalability.
Wei, Dong, "Optimization opportunities in human in the loop computational paradigm" (2022). Dissertations. 1612.