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

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Yixuan, Li, Zhiyuan, He, Weijia, Dai, Mian, Shen, Chen, Long, Teng, Zhang, Hongbin
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918321094393856
author Zhang, Yixuan
Li, Zhiyuan
He, Weijia
Dai, Mian
Shen, Chen
Long, Teng
Zhang, Hongbin
author_facet Zhang, Yixuan
Li, Zhiyuan
He, Weijia
Dai, Mian
Shen, Chen
Long, Teng
Zhang, Hongbin
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2602_03319
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Information-Theoretic Multi-Model Fusion for Target-Oriented Adaptive Sampling in Materials Design
Zhang, Yixuan
Li, Zhiyuan
He, Weijia
Dai, Mian
Shen, Chen
Long, Teng
Zhang, Hongbin
Machine Learning
Materials Science
Information Theory
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.
title Information-Theoretic Multi-Model Fusion for Target-Oriented Adaptive Sampling in Materials Design
topic Machine Learning
Materials Science
Information Theory
url https://arxiv.org/abs/2602.03319