Towards Reliable AI: Adequacy Metrics for Ensuring the Quality of AI-based Systems
Anonymous · Zenodo (CERN European Organization for Nuclear Research) · 2023
This repository contains dataset required for the replication of results publishes in the article "Towards Reliable AI: Adequacy Metrics for Ensuring the Quality of AI-based Systems". The detail of the directories and files included in this dataset are 1: Testsuite1: contains test suite generated for testing of Baidu Apollo (in CSV format). 2. Testsuite2: contains test suite generate for testing of AV using AsFault and BeamNG (in CSV format) 3. ISA-code: contains code for generation of TISA metrics. This code is written in Matlab, and to run this, a shell script "run-ISA-matlab.sh" is provided. For recording the TISA metrics, a csv is required in the "results" directory. The format of the CSV is also provided in this directory. 4. Diversity metrics: contains python implementation of the diversity metrics used in our study.