With machines, Artificial Intelligence, and Machine Learning becoming increasingly ubiquitous in our society, we need to start thinking about the implications and ethical concerns of new models. A data scientist can view the world of unfair or biased models in three layers which include social injustice bias and measurement bias. I propose that we provide two different types of solutions for the two types of bias. Social injustice bias found in machine learning models could be fought by using diversity constraints. Measurement bias, on the other hand, could be fought by using fairness constraints. By looking at the application of hiring, where both social injustice bias and measurement bias is prolific, I define the hiring (or cohort selection) problem as a Multi-Armed Bandit problem. I will discuss the theoretical results of the defined MAB problem as well as discuss a few open problems in both fairness and diversity. Finally I will suggest broader impact applications for both of these domains.

Examining Committee: 

 

                          Chair:               Dr. John Dickerson
                          Dept. rep:        Dr. Jeff Foster
                          Members:        Dr. Hal Daumé