Can We Just Get Along? The Role of Human-In-The-Loop Machine Learning Protocol for Advancing Flood Evacuation Decision Making
Abstract
Machine learning simulators cannot fulfill their decision-support potential on their own. Rather, they must be utilized in partnership with the human to co-create collective intelligence. Human-in-the-loop machine learning (HITL ML) allows for the integration of the best of human intelligence with the best of machine intelligence. The effectiveness of HITL ML partnership requires much more than old-school manual simulation that consisted mainly of running an ML simulator and feeding the data. Here a framework for the human in the loop is discussed that provides a systematic approach to engineer potential solutions for flood evacuation decision-making. The barriers and motivations for applicability of the HITL ML teaming are also explored. This presentation discusses a case study of a HITL ML teaming for a flood evacuation problem with elements of lessons learned that can be transferred to other water resources decision-making situations. This research is funded by the US National Science Foundation.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H53E..07S