Guaranteed Approximation Bounds for Mixed-Precision Neural Operators.
Renbo TuColin WhiteJean KossaifiBoris BonevGennady PekhimenkoKamyar AzizzadenesheliAnima AnandkumarPublished in: ICLR (2024)
Keyphrases
- error tolerance
- error bounds
- stage stochastic programs
- network architecture
- neural network
- approximation methods
- upper bound
- lower bound
- precision and recall
- worst case
- high precision
- approximation algorithms
- closed form
- tight bounds
- bio inspired
- relative error
- building blocks
- lower and upper bounds
- morphological operators
- average case
- neural model
- approximation error
- theoretical analysis
- high recall
- spike trains
- data sets
- genetic algorithm
- neural fuzzy
- biologically inspired