Published June 1, 2021 | Version v1
Report Open

Uncertainty-Wizard: Fast and User-Friendly Neural Network Uncertainty Quantification

  • 1. Università della Svizzera italiana

Description

Uncertainty and confidence have been shown to be useful metrics in a wide variety of techniques proposed for deep learning testing, including test data selection and system supervision. We present Uncertainty-Wizard, a tool that allows to quantify such uncertainty and confidence in  artificial neural networks. It is built on top of the industry-leading tf.keras deep learning API and it provides a near-transparent and  easy to understand interface. At the same time, it includes major performance optimizations that we benchmarked on two different machines and different configurations.

Files

TR-Precrime-2021-04.pdf

Files (658.1 kB)

Name Size Download all
md5:4f6500a64a0b1d256f212f8fb2a44e68
658.1 kB Preview Download

Additional details

Related works

Is obsoleted by
10.1109/ICST49551.2021.00056 (DOI)

Funding

PRECRIME – Self-assessment Oracles for Anticipatory Testing 787703
European Commission