Posterior Covariance

Posterior covariance estimation focuses on quantifying uncertainty in probabilistic models, particularly in the context of Bayesian inference. Current research emphasizes efficient computation of posterior covariance, exploring methods within Gaussian processes and neural networks, often leveraging techniques like linear response and maximum likelihood estimation to improve accuracy and scalability. This improved understanding of uncertainty is crucial for enhancing the reliability of predictions in various applications, including inverse problems, speaker verification, and generally improving the robustness of machine learning models. The development of accurate and computationally feasible posterior covariance estimation methods is driving progress across numerous fields.

Papers

November 1, 2024

pcaGAN: Improving Posterior-Sampling cGANs via Principal Component Regularization
Matthew C. Bendel, Rizwan Ahmad, Philip Schniter
Inverse Problem Regularization Model Large Relevance Improvement Posterior Sampling Principal Component Posterior Covariance

October 29, 2024

Unpicking Data at the Seams: VAEs, Disentanglement and Independent Components
Carl Allen
Variational Autoencoder Variational Autoencoders Disentanglement Capability Connected Component Generative Paradigm Posterior Covariance

September 6, 2024

Epistemic Uncertainty and Observation Noise with the Neural Tangent Kernel
Sergio Calvo-Ordoñez, Konstantina Palla, Kamil Ciosek
Gradient Descent Neural Tangent Kernel Epistemic Uncertainty Wide Neural Network Noisy Observation Posterior Covariance

August 14, 2024

Posterior Covariance Structures in Gaussian Processes
Difeng Cai, Edmond Chow, Yuanzhe Xi
Gaussian Process Posterior Covariance

February 3, 2024

Improving Diffusion Models for Inverse Problems Using Optimal Posterior Covariance
Xinyu Peng, Ziyang Zheng, Wenrui Dai, Nuoqian Xiao, Chenglin Li, Junni Zou, Hongkai Xiong
Diffusion Model Inverse Problem Covariance Conditioning Posterior Covariance

April 11, 2023

Black Box Variational Inference with a Deterministic Objective: Faster, More Accurate, and Even More Black Box
Ryan Giordano, Martin Ingram, Tamara Broderick
Variational Inference Black Box Deterministic Algorithm Posterior Approximation Variational Approximation Black Box Variational Inference Mean Field Variational Posterior Covariance

February 23, 2023

Incorporating Uncertainty from Speaker Embedding Estimation to Speaker Verification
Qiongqiong Wang, Kong Aik Lee, Tianchi Liu
High Uncertainty Anticipation Estimation Task Speaker Verification Discriminant Analysis Uncertainty Propagation Speech Driven Posterior Covariance

December 16, 2022

Efficient Conditionally Invariant Representation Learning
Roman Pogodin, Namrata Deka, Yazhe Li, Danica J. Sutherland, Victor Veitch, Arthur Gretton
Conditional Independence Ridge Regression Domain Invariant Dependent Feature Posterior Covariance Conditional Invariant

May 30, 2022

Posterior and Computational Uncertainty in Gaussian Processes
Jonathan Wenger, Geoff Pleiss, Marvin Pförtner, Philipp Hennig, John P. Cunningham
Gaussian Process Covariance Conditioning Posterior Mode Computational Uncertainty Posterior Covariance