Estimating the Brittleness of AI: Safety Integrity Levels and the Need for Testing Out-Of-Distribution Performance.
Andrew J. LohnPublished in: CoRR (2020)
Keyphrases
- artificial intelligence
- expert systems
- case based reasoning
- data distribution
- knowledge representation
- ai systems
- gaussian distribution
- random variables
- test set
- intelligent systems
- case study
- ai methods
- real time
- power law
- lower levels
- levels of abstraction
- ai technologies
- spatial distribution
- test data
- software engineering
- object oriented
- machine learning
- neural network
- data sets