Interpretable evolutionary algorithm improves multiobjective decision-making
An Interactive Evolutionary Multiobjective Optimization Algorithm Based on Interpretable Active Preference Learning
Neural and Evolutionary ComputingArtificial Intelligence
Summary
Many decisions involve balancing multiple goals at once. The authors developed a new evolutionary algorithm that learns what a person prefers using simple, understandable rules. Unlike other methods that act like 'black boxes,' this one shows why certain choices are made, helping people trust the results. Tests show it finds the best solutions people want more effectively, even if it doesn't perform as well overall. The method also helps explain which factors matter most in the decision.
What this means in practice
- •For engineering design teams: Use the method to guide product or system designs by learning and explaining decision-makers’ preferences from comparisons.
- •For healthcare decision teams: Apply the algorithm for transparent multi-objective treatment planning where explanations support accountability.
Authors
Thomas Quadt, Boris Čule, Roy Lindelauf, Herman Monsuur, Mark Voskuijl
Abstract
Many decision-making tasks are inherently multi-objective. Interactive evolutionary multi-objective optimization algorithms (iEMOAs) are particularly well-suited to help the decision-maker find their most preferred solution. However, many of these algorithms employ black-box preference models. For high-stakes decision-making tasks, like healthcare and defense, a high level of transparency is needed, for instance to support accountability. We propose an interpretable iEMOA which uses active learning to efficiently learn a tree-based preference model directly from pairwise comparisons. The survival operator of NSGA-II and NSGA-III is extended with the learned preference model to guide the algorithm towards the region of interest. An extensive experimental study compares the proposed method to a state-of-the-art black-box model and a state-of-the-art interpretable model. Results indicate that our proposed model outperforms the alternatives in terms of finding the most preferred solution at the cost of a slightly worse average final population. Finally, a case study on a real-life engineering problem illustrates how inspecting the learned model reveals which objectives drive the decision-maker's preferences.