A General Framework for quantifying Aleatoric and Epistemic uncertainty in Graph Neural Networks

05/20/2022
by   Sai Munikoti, et al.
0

Graph Neural Networks (GNN) provide a powerful framework that elegantly integrates Graph theory with Machine learning for modeling and analysis of networked data. We consider the problem of quantifying the uncertainty in predictions of GNN stemming from modeling errors and measurement uncertainty. We consider aleatoric uncertainty in the form of probabilistic links and noise in feature vector of nodes, while epistemic uncertainty is incorporated via a probability distribution over the model parameters. We propose a unified approach to treat both sources of uncertainty in a Bayesian framework, where Assumed Density Filtering is used to quantify aleatoric uncertainty and Monte Carlo dropout captures uncertainty in model parameters. Finally, the two sources of uncertainty are aggregated to estimate the total uncertainty in predictions of a GNN. Results in the real-world datasets demonstrate that the Bayesian model performs at par with a frequentist model and provides additional information about predictions uncertainty that are sensitive to uncertainties in the data and model.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
05/15/2023

Integrating Uncertainty into Neural Network-based Speech Enhancement

Supervised masking approaches in the time-frequency domain aim to employ...
research
09/20/2019

Characterizing Sources of Uncertainty to Proxy Calibration and Disambiguate Annotator and Data Bias

Supporting model interpretability for complex phenomena where annotators...
research
12/26/2020

Bayesian Inductive Learner for Graph Resiliency under uncertainty

In the quest to improve efficiency, interdependence and complexity are b...
research
08/01/2023

Capsa: A Unified Framework for Quantifying Risk in Deep Neural Networks

The modern pervasiveness of large-scale deep neural networks (NNs) is dr...
research
11/17/2020

Quantifying Uncertainty from Different Sources in Deep Neural Networks for Image Classification

Quantifying uncertainty in a model's predictions is important as it enab...
research
04/20/2021

A Bayesian Convolutional Neural Network for Robust Galaxy Ellipticity Regression

Cosmic shear estimation is an essential scientific goal for large galaxy...
research
09/05/2021

Stochastic Neural Radiance Fields:Quantifying Uncertainty in Implicit 3D Representations

Neural Radiance Fields (NeRF) has become a popular framework for learnin...

Please sign up or login with your details

Forgot password? Click here to reset