T-DominO

Abstract Real-world design problems are a messy combination of constraints, objectives, and features. Exploring these problem spaces can be defined as Multi-Criteria Exploration (MCX) problem, whose goals to produce set diverse solutions with high performance across many while avoiding low any objectives. Quality-Diversity algorithms the needed variation, but typically consider only single objective. We present new ranking, T-DominO, specifically designed handle multiple objectives in MCX problems. T-DominO ranks individuals relative other archive, favoring balanced over those which excel at few cost others. Keeping solution each MAP-Elites bin maintains visual accessibility archive – strong asset for exploration. illustrate our approach on easily understood benchmarks, showcase its potential many-objective real-world architecture case study.