Skip to content

Latest commit

 

History

History
13 lines (10 loc) · 489 Bytes

README.md

File metadata and controls

13 lines (10 loc) · 489 Bytes

vnncomp2022

Benchmark problem from Collins Aerospace for the 2022 VNN Competition

The submission includes:

  • data folder: data files (.mat) used to generate properties
  • onnx folder: trained neural networks
  • specs folder: generated vnnlib specifications
  • benchmark description (PDF file)
  • description of instances with respective timeouts (CSV file)
  • python script to generate properties, parameterized with random seed

For details, see benchmark_description.pdf.