A collaborative interactive genetic algorithm framework for mixed-initiative interaction with human and simulated experts: A case study in long-term groundwater monitoring design

The Interactive Genetic Algorithm (IGA) allows water resources and environmental decision makers to become active online participants during the optimization process, and thus provides a method to include qualitative expert knowledge within the search criteria. However, various interfering human factors, especially human fatigue, can limit the extent of the decision maker's participation. In this paper, we propose a mixed-initiative interaction technique for the IGA in which a simulated expert (created by using a machine learning model) can share the workload of interaction with the human expert, while constantly learning her/his preferences. This collaborative framework also allows the system to observe the learning behaviors of both the human and simulated expert, while utilizing their knowledge for search purposes. Many machine learning models can be utilized for creating the simulated experts, in our work we use fuzzy logic modeling that implements a rule based decision making criteria for modeling the human expert's preferences. These methodologies are tested on a field scale groundwater monitoring application to analyze their benefits.