PulseAugur
中
实时 13:58:36
English(EN) Accurate Large-sample Uncertainty Quantification using Stochastic Gradient Markov Chain Monte Carlo

新方法增强了机器学习模型的不确定性量化

作者 PulseAugur 编辑部 · [5 个来源] ·

研究人员开发了新的理论框架和实用方法来改进机器学习模型的不确定性量化。一篇论文介绍了随机梯度方法的离散时间近似，以准确估计协方差和自相关时间，为大批量或错误指定模型提供更好的调优指导。另一项研究提出了SGLD-Gibbs的统计缩放极限理论，为潜在变量模型提供原则性的超参数调优，从而获得更有意义的不确定性估计。第三篇论文提出了一种基于高斯过程的方法，用于因果干预函数的不确定性量化，提高了高风险应用的可靠性。 AI

影响 不确定性量化方面的这些进展可能带来更可靠、更值得信赖的AI系统，特别是在理解模型置信度至关重要的关键应用中。

排序理由 该集群包含多篇在arXiv上发表的学术论文，详细介绍了机器学习方面新的理论和方法进展，特别是关于不确定性量化。

在 arXiv stat.ML 阅读 →

AI 生成摘要 · Google Gemini · 来自 5 个来源。 我们如何撰写摘要 →

报道来源 [5]

  1. arXiv stat.ML TIER_1 English(EN) · Yu Wang, Jie Ding, Jonathan H. Huggins ·

    使用随机梯度马尔可夫链蒙特卡洛进行精确的大样本不确定性量化

    arXiv:2606.00293v1 Announce Type: cross Abstract: Tuning algorithms such as stochastic gradient descent (SGD) and stochastic gradient Langevin dynamics (SGLD) for approximate sampling and uncertainty quantification remains challenging, particularly in the practically relevant set…

  2. arXiv stat.ML TIER_1 English(EN) · Xiaoyu Wang, Jonathan H. Huggins ·

    使用子采样马尔可夫链蒙特卡洛方法对潜在变量模型进行大规模不确定性量化

    arXiv:2606.00309v1 Announce Type: cross Abstract: Stochastic gradient Langevin dynamics combined with Gibbs updates (SGLD--Gibbs) provides a highly scalable approach to approximate Bayesian inference in latent variable models. However, it remains unclear how to tune the algorithm…

  3. arXiv stat.ML TIER_1 English(EN) · Hugh Dance, Peter Orbanz, Arthur Gretton ·

    因果不确定性量化的干预过程

    arXiv:2410.14483v3 Announce Type: replace Abstract: Reliable uncertainty quantification for causal effects is crucial in high-stakes applications, but remains challenging when the target is an entire function rather than a scalar estimand. In this work, we introduce a GP-based ap…

  4. arXiv stat.ML TIER_1 English(EN) · Jonathan H. Huggins ·

    Large-scale Uncertainty Quantification for Latent Variable Models Using Subsampling Markov Chain Monte Carlo

    Stochastic gradient Langevin dynamics combined with Gibbs updates (SGLD--Gibbs) provides a highly scalable approach to approximate Bayesian inference in latent variable models. However, it remains unclear how to tune the algorithm's hyperparameters in a principled manner to ensur…

  5. arXiv stat.ML TIER_1 English(EN) · Jonathan H. Huggins ·

    Accurate Large-sample Uncertainty Quantification using Stochastic Gradient Markov Chain Monte Carlo

    Tuning algorithms such as stochastic gradient descent (SGD) and stochastic gradient Langevin dynamics (SGLD) for approximate sampling and uncertainty quantification remains challenging, particularly in the practically relevant settings when the batch size is large or the model is…

相关实体

相关话题