Information-Theoretic Multi-Model Fusion for Target-Oriented Adaptive Sampling in Materials Design

2026-02-03

AI Summary

This work addresses the challenge of high-dimensional heterogeneous material design by reframing goal-oriented discovery as a trajectory optimization problem, wherein low-entropy information states are maintained to focus exploration on target-relevant regions. The authors propose an information-theoretic multi-model fusion framework that integrates data, model beliefs, and physical or structural priors. Key innovations include a dimension-aware information budget, adaptive distillation of heterogeneous surrogate models, and structure-aware candidate manifold analysis, which jointly harmonize consensus-driven exploitation with discrepancy-driven exploration. Evaluated across 14 single- and multi-objective tasks, the approach demonstrates markedly improved sample efficiency and reliability, consistently reaching high-performance regions within approximately one hundred evaluations. Its robustness to complex response landscapes is further validated on a 20-dimensional synthetic benchmark.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Mixed Discrete/Continuous Optimization
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Target-oriented discovery under limited evaluation budgets requires making reliable progress in high-dimensional, heterogeneous design spaces where each new measurement is costly, whether experimental or high-fidelity simulation. We present an information-theoretic framework for target-oriented adaptive sampling that reframes optimization as trajectory discovery: instead of approximating the full response surface, the method maintains and refines a low-entropy information state that concentrates search on target-relevant directions. The approach couples data, model beliefs, and physics/structure priors through dimension-aware information budgeting, adaptive bootstrapped distillation over a heterogeneous surrogate reservoir, and structure-aware candidate manifold analysis with Kalman-inspired multi-model fusion to balance consensus-driven exploitation and disagreement-driven exploration. Evaluated under a single unified protocol without dataset-specific tuning, the framework improves sample efficiency and reliability across 14 single- and multi-objective materials design tasks spanning candidate pools from $600$ to $4 \times 10^6$ and feature dimensions from $10$ to $10^3$, typically reaching top-performing regions within 100 evaluations. Complementary 20-dimensional synthetic benchmarks (Ackley, Rastrigin, Schwefel) further demonstrate robustness to rugged and multimodal landscapes.

Problem

• target-oriented discovery
• limited evaluation budgets
• high-dimensional design spaces
• adaptive sampling
• materials design

Innovation

• information-theoretic fusion
• target-oriented adaptive sampling
• multi-model fusion
• heterogeneous surrogate reservoir
• low-entropy information state