print · login   

Random Moore machines

Tag: random


We have included two sets of randomly generated Moore machines in our repository:

  • $3\,600$ Moore machiness with $1\,000$ states each over the alphabet $\Sigma = \{ 0, 1, \ldots, 19\}$ and output alphabet $\Gamma = \Sigma$; and
  • $3\,600$ Moore machines with $2\,000$ states each over the alphabet $\Sigma = \{ 0, 1, \ldots, 9\}$ and output alphabet $\Gamma = \Sigma$.

The number of states and the number of elements in the input/output alphabets of these automata were chosen to be challenging, though still manageable for state-of-the-art learning and testing algorithms. To generate Moore machines, we used the following two-step process: first, we randomly generated a DFA using libalf's off-the-shelf implementation of Champarnaud and Parantho├źn's method; second, we assigned to each state an output symbol that was drawn uniformly at random from the output alphabet.