English(EN) Improving the Accuracy of Amortized Model Comparison with Self-Consistency

自洽性损失提高了贝叶斯模型比较的准确性

作者 PulseAugur 编辑部 · [1 个来源] · 2026-05-13 04:00

研究人员开发了一种自洽性（SC）损失，用于在模拟模型被错误指定时提高摊销贝叶斯模型比较（BMC）的准确性。该技术增强了BMC估计器，特别是在所有候选模型都不完美的开放世界场景中。研究评估了四种摊销BMC方法，发现即使在模型被错误指定的情况下，当解析似然可用或代理似然局部准确时，SC训练也能显著提高性能。 AI

影响增强了用于训练和评估机器学习模型的统计方法。

排序理由详细介绍统计模型比较新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv stat.ML 阅读 →

Šimon Kucharský

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

报道来源 [1]

arXiv stat.ML TIER_1 English(EN) · \v{S}imon Kucharsk\'y, Aayush Mishra, Daniel Habermann, Stefan T. Radev, Paul-Christian B\"urkner · 2026-05-13 04:00

Improving the Accuracy of Amortized Model Comparison with Self-Consistency

arXiv:2508.20614v3 Announce Type: replace Abstract: Amortized Bayesian model comparison (BMC) enables fast probabilistic ranking of models via simulation-based training of neural surrogates. However, the accuracy of neural surrogates deteriorates when simulation models are misspe…

报道来源 [1]

Improving the Accuracy of Amortized Model Comparison with Self-Consistency

相关话题