New framework enables precise passive learning for expressive timed logic

Researchers have developed a novel framework for the precise passive learning of Metric Interval Temporal Logic (MITL), an expressive timed logic crucial for verifying real-time systems. This approach tackles the challenge of learning MITL without predefined templates or restricted logic fragments by reducing the timed learning problem to a scalable untimed one. The method identifies quantitative timing differences between positive and negative traces to synthesize precise timed constraints, which are then integrated as new Boolean atomic propositions. This technique allows for the delegation of complex formula evaluation to optimized, off-the-shelf untimed LTL tools, and the framework is proven to be complete, guaranteeing that a separating specification can always be found. AI

IMPACT This research could advance the verification of real-time systems by enabling more precise automated specification mining.

RANK_REASON The item is a research paper published on arXiv detailing a new framework for a specific type of logic. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• Metric Interval Temporal Logic
• ScienceCast
• Shankara Narayanan Krishna
• Timed Regular Expressions

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →