Calibrated Uncertainty for Molecular Property Prediction using Ensembles of Message Passing Neural Networks.
Jonas BuskPeter Bjørn JørgensenArghya BhowmikMikkel N. SchmidtOle WintherTejs VeggePublished in: CoRR (2021)
Keyphrases
- message passing
- neural network ensemble
- neural network
- belief propagation
- probabilistic inference
- distributed systems
- competitive learning
- factor graphs
- back propagation
- graphical models
- shared memory
- prediction accuracy
- approximate inference
- inference in graphical models
- markov random field
- sum product algorithm
- sum product
- distributed shared memory
- ldpc codes
- graph cuts
- multi view
- probabilistic graphical models
- junction tree
- matrix multiplication
- artificial neural networks
- three dimensional