QUIVER: Cost-Aware Adaptive Preference Querying in Surrogate-Assisted Evolutionary Multi-Objective Optimization
arXiv:2605.04267v2 Announce Type: replace Abstract: Interactive multi-objective optimization systems face a budget allocation dilemma: one can spend resources on expensive objective evaluations or on eliciting decision-maker preferences that identify the relevant region of the Pareto set. Moreover, preference elicitation itself spans modalities with different information content and cognitive burden, ranging from cheap, noisy pairwise preference statements (PS) to richer but costlier indifference adjustments (IA). We study cost-aware optimization under an unknown scalarization and introduce QU
This development appears now as the field of AI and multi-objective optimization matures, seeking more efficient and cost-effective ways to integrate human preferences into complex decision-making systems.
It is important because it addresses a core challenge in interactive AI: balancing computational cost with the quality of human input, which is crucial for practical applications of AI in real-world scenarios.
This research could lead to more adaptive and efficient AI systems that better understand and incorporate user preferences without excessive resource expenditure or cognitive burden on users.
- · AI developers
- · Optimization researchers
- · Industries using interactive AI
- · Inefficient preference elicitation methods
Improved decision-making capabilities in AI systems that require human input for preference articulation.
Reduced operational costs and faster deployment of AI solutions across various sectors due to more efficient preference learning.
Enhanced user trust and adoption of AI systems as they become more aligned with human values and objectives through better preference integration.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG