New framework learns uncertain temporal logic specifications from system demonstrations

Researchers have developed a new framework for learning Linear Temporal Logic (LTL) formulas from system demonstrations that may contain uncertainty. This method addresses practical challenges where system traces can be incomplete or erroneous due to issues like sensor faults or data loss. By modeling uncertainty using Hamming distance and generating estimates around observed traces, the framework reduces the problem to Pseudo-Boolean Optimization, aiming to recover specifications that more accurately reflect ground truth under uncertain conditions. AI

IMPACT Enhances formal verification and controller synthesis in safety-critical systems by improving the robustness of learning from uncertain data.

RANK_REASON Academic paper detailing a new framework for learning temporal logic specifications. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Hamming distance
• Linear Temporal Logic
• Pseudo-Boolean Optimization

• paper
• other

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