Date: October 2nd–3rd, 2023
Venue:
National Institute of Informatics,
2-1-2 Hitotsubashi, Chiyoda-ku, Tokyo

October 2nd

• 13:00-13:30: Opening and project status (Kohei)
• 13:30-14:30: Learning nonlinear hybrid automata from input–output time-series data (Amit) https://doi.org/10.1007/978-3-031-45329-8_2
• 14:30-15:00: Break
• 15:00-16:00: Perception Contracts for Safety of ML-Enabled Systems (Chiao) https://doi.org/10.1145/3622875
• 16:00-16:30: Break
• 16:30-17:30: Compositional Probabilistic Model Checking with String Diagrams of MDPs. (Ichiro, invited from ERATO),
  • Compositional MDP model checking https://doi.org/10.1007/978-3-031-37709-9_3
  • Kittiphon’s algorithm [Phalakarn, Takisaka, Haas & Hasuo, CAV’20] https://doi.org/10.1007/978-3-030-53291-8_19

October 3rd

• 10:00-11:00: Repair Techniques for Deep Neural Networks for Safety in Perception of Driving Systems (Paolo, invited from eAI project)
  • References
    • M. Duran, X. Zhang, P. Arcaini, F. Ishikawa. What to Blame? On the Granularity of Fault Localization for Deep Neural Networks.
    • D. Li Calsi, M. Duran, X. Zhang, P. Arcaini, F. Ishikawa. Distributed Repair of Deep Neural Networks. ICST’23
    • D. Li Calsi, M. Duran, T. Laurent, X. Zhang, P. Arcaini, F. Ishikawa. Adaptive Search-based Repair of Deep Neural Networks. GECCO’23
• 11:00-11:30: Break
• 11:30-12:30: Probabilistic Black-Box Checking via Active MDP Learning (Kohei)
  • Replay of EMSOFT 2023 work Probabilistic Black-Box Checking via Active MDP Learning | ACM Transactions on Embedded Computing Systems https://doi.org/10.1145/3609127
• 12:30-14:00: Lunch
• 14:00-15:00: Chance Constrained Probability Measure Optimization (Shen Xun: Assistant Prof. from Osaka University, only Oct. 3 seems available)
  • https://arxiv.org/abs/2309.08231
• 15:00-15:30: Break
• 15:30-16:30: WiP: Oblivious Online Monitoring for STL with Arithmetic Predicates via Fully Homomorphic Encryption (Masaki)
  • https://doi.org/10.1007/978-3-031-13185-1_22
• 16:30-18:00: Discussion