Output-Constrained Bayesian Neural Networks

05/15/2019
by   Wanqian Yang, et al.
0

Bayesian neural network (BNN) priors are defined in parameter space, making it hard to encode prior knowledge expressed in function space. We formulate a prior that incorporates functional constraints about what the output can or cannot be in regions of the input space. Output-Constrained BNNs (OC-BNN) represent an interpretable approach of enforcing a range of constraints, fully consistent with the Bayesian framework and amenable to black-box inference. We demonstrate how OC-BNNs improve model robustness and prevent the prediction of infeasible outputs in two real-world applications of healthcare and robotics.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
10/21/2020

Incorporating Interpretable Output Constraints in Bayesian Neural Networks

Domains where supervised models are deployed often come with task-specif...
research
07/14/2021

Hybrid Bayesian Neural Networks with Functional Probabilistic Layers

Bayesian neural networks provide a direct and natural way to extend stan...
research
06/15/2023

Neural Fields with Hard Constraints of Arbitrary Differential Order

While deep learning techniques have become extremely popular for solving...
research
06/18/2020

Predictive Complexity Priors

Specifying a Bayesian prior is notoriously difficult for complex models ...
research
07/08/2023

On Regularization and Inference with Label Constraints

Prior knowledge and symbolic rules in machine learning are often express...
research
07/26/2017

Enforcing Constraints on Outputs with Unconstrained Inference

Increasingly, practitioners apply neural networks to complex problems in...
research
03/23/2020

Sample-Specific Output Constraints for Neural Networks

Neural networks reach state-of-the-art performance in a variety of learn...

Please sign up or login with your details

Forgot password? Click here to reset