Deep Ensembles
PulseAugur coverage of Deep Ensembles — every cluster mentioning Deep Ensembles across labs, papers, and developer communities, ranked by signal.
3 天有情绪数据
-
深度集成未能捕捉图神经网络中的不确定性
一项新的研究论文质疑了深度集成在图神经网络中进行不确定性量化的有效性。研究发现,集成模型相比单一模型在不确定性量化方面几乎没有改进,其收益主要来自稳定预测而非提高不确定性估计。这归因于“认知崩溃”,即独立训练的网络产生过于相似的预测,从而抵消了集成模型的核心优势。
-
New framework unifies uncertainty-aware explainable AI
Researchers have introduced a new framework for explainable AI (XAI) that incorporates uncertainty awareness, moving beyond deterministic attribution maps. This approach formalizes the 'explanation distribution' derived…
-
研究论文区分了用于人工智能不确定性的交叉验证与深度集成
一篇题为“折叠中的迷失”的新研究论文强调了人工智能研究中关于医学图像分割不确定性估计的一个普遍误解。研究表明,使用K折交叉验证(CV)来形成集成模型,通常被错误地标记为深度集成(DE),这可能导致对不确定性的不准确解读。研究发现,使用相同训练数据但不同随机种子的DE更适合故障检测等可靠性任务,而CV集成模型更适合建模模糊性。
-
AI models show improved blood pressure estimation reliability
Researchers investigated the reliability of uncertainty quantification in deep learning models for blood pressure estimation from photoplethysmography (PPG) signals. The study found that deep ensembles (DE) offer greate…
-
Singular Bayesian Neural Networks
研究人员推出了一种名为Singular Bayesian Neural Networks的新方法,该方法显著减少了贝叶斯神经网络所需的参数数量。通过使用低秩分解来参数化权重,这些网络将其后验集中在秩流形上,与标准的均值场方法相比,能够更有效地进行相关性建模。该技术提供了改进的泛化界限和具有竞争力的预测性能,实证结果显示参数数量减少高达33倍,并且增强了分布外检测能力。
-
Bayesian deep learning evaluation unstable in low-data settings, studies find
Two new arXiv papers highlight significant instability in evaluating Bayesian deep learning methods, particularly under data scarcity. Researchers found that standard evaluation metrics can produce unreliable and datase…