Laplace Approximation
PulseAugur coverage of Laplace Approximation — every cluster mentioning Laplace Approximation across labs, papers, and developer communities, ranked by signal.
-
新的SSLA方法改进了贝叶斯模型不确定性量化
研究人员开发了一种名为自监督拉普拉斯近似（SSLA）的新方法，用于直接近似贝叶斯模型中的后验预测分布。该方法借鉴了自训练技术的灵感，通过在模型自身预测上重新拟合模型来量化预测不确定性。SSLA方法提供了一种确定性的、无采样的近似方法，在回归任务（包括贝叶斯神经网络）的预测校准方面优于经典的拉普拉斯近似，同时保持了计算效率。
-
New methods improve Laplace approximation for neural network uncertainty
Researchers have developed new methods for approximating the Laplace approximation in deep neural networks, addressing the computational challenges of inverting large Hessian matrices. The proposed Gradient-Laplace and …
-
Bayesian Tensor Network Kernel Machines use Laplace approximation for uncertainty estimation
Researchers have developed a new Bayesian Tensor Network Kernel Machine (LA-TNKM) that utilizes a linearized Laplace approximation for inference. This method addresses the challenge of providing uncertainty estimates in…