用狄利克雷分布改进随机丢弃，让神经网络更准确地估计不确定性。
Dirichlet-Based Monte Carlo Dropout for Uncertainty Estimation in Neural Networks
- 在随机丢弃基础上引入狄利克雷分布建模分类概率
- 提升不确定性估计的校准度，计算开销与原方法相当
- 适合需要可信置信度的场景，如医疗诊断和自动驾驶
传统神经网络提供确定性输出，缺乏内在的不确定性估计。虽然贝叶斯神经网络（BNN）能从理论上量化不确定性，但其计算复杂度限制了可扩展性。蒙特卡洛（MC）丢弃最初作为正则化技术提出，已被证明可通过多次随机前向传播近似贝叶斯推断。本文在MC丢弃中引入狄利克雷分布框架，借鉴Sensoy等（2018）提出的分类概率建模方式，使不确定性表示更具信息量。该方法在保持MC丢弃计算高效性的前提下，显著提升了不确定性估计质量。我们分析了方法的理论基础，并与现有技术进行对比，实验表明该方法能生成更校准的不确定性估计，为不确定感知深度学习模型提供了一种实用解决方案。
原文摘要 · Abstract (English)
Traditional neural networks provide deterministic predictions without inherent uncertainty estimates. While Bayesian Neural Networks (BNNs) offer a principled approach to uncertainty quantification, their computational complexity limits scalability. Monte Carlo (MC) Dropout, initially introduced as a regularization technique, has been shown to approximate Bayesian inference by enabling probabilistic modeling through multiple stochastic forward passes. In this work, we enhance uncertainty estimation in deep learning by integrating a Dirichlet-based framework within MC Dropout. Specifically, we leverage the formulation proposed by Sensoy et al. (2018), where class probabilities are modeled using a Dirichlet distribution, allowing for a more informative uncertainty representation. The proposed approach maintains the computational efficiency of MC Dropout while improving the quality of uncertainty estimates. We discuss the theoretical foundations of our method and compare it with existing uncertainty quantification techniques. The results highlight the effectiveness of the proposed method in producing well-calibrated uncertainty estimates, offering a practical solution for uncertainty-aware deep learning models.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。