Doctoral Thesis: Systems for Usable Machine Learning

Wednesday, October 9
9:00 am - 10:30 am

45-600B

By: Alexandra (Ola) Zytek

Thesis Supervisor: Kalyan Veeramachaneni

Details

  • Date: Wednesday, October 9
  • Time: 9:00 am - 10:30 am
  • Category:
  • Location: 45-600B
Additional Location Details:

Abstract: Many real-world decision problems are complex, with outcomes difficult to measure and evaluate. Individual mistakes can lead to significant costs, and computational tools such as ML models must be integrated alongside existing, well-established human workflows. These properties of such decision problems means that ML solutions must be usable in order to be effective — in other words, developed and deployed in such a way as to be used by humans in decision-making and improve outcomes.

In this thesis, I use real-world case studies to synthesize generalizable lessons for applying usable ML tools to complex, real-world decision problems. Based on experience developing ML tools for child welfare screening, I discuss the design and development of several systems to generate understandable, effective ML explanations and tools. I then discuss our case study in applying these systems to the decision problem of wind turbine monitoring.

I finish by discussing the practical lessons from this work for future development of usable ML, and the remaining challenges in applying ML to such complex real-world domains.