Towards Efficient Active Learning of PDFA
This work addresses the problem of improving efficiency in active learning for PDFA, which is incremental as it builds on existing methods with specific enhancements.
The paper tackled the problem of active learning for PDFA by introducing a new algorithm that incorporates a state congruence based on next-symbol probability distributions, quantization to handle distribution differences, and an efficient tree-based data structure, resulting in significant performance gains compared to reference implementations.
We propose a new active learning algorithm for PDFA based on three main aspects: a congruence over states which takes into account next-symbol probability distributions, a quantization that copes with differences in distributions, and an efficient tree-based data structure. Experiments showed significant performance gains with respect to reference implementations.