Abstract:Temporal reasoning with conditionals is more complex than both classical temporal reasoning and reasoning with timeless conditionals, and can lead to some rather counter-intuitive conclusions. For instance, Aristotle's famous "Sea Battle Tomorrow" puzzle leads to a fatalistic conclusion: whether there will be a sea battle tomorrow or not, but that is necessarily the case now. We propose a branching-time logic LTC to formalise reasoning about temporal conditionals and provide that logic with adequate formal semantics. The logic LTC extends the Nexttime fragment of CTL*, with operators for model updates, restricting the domain to only future moments where antecedent is still possible to satisfy. We provide formal semantics for these operators that implements the restrictor interpretation of antecedents of temporalized conditionals, by suitably restricting the domain of discourse. As a motivating example, we demonstrate that a naturally formalised in our logic version of the `Sea Battle' argument renders it unsound, thereby providing a solution to the problem with fatalist conclusion that it entails, because its underlying reasoning per cases argument no longer applies when these cases are treated not as material implications but as temporal conditionals. On the technical side, we analyze the semantics of LTC and provide a series of reductions of LTC-formulae, first recursively eliminating the dynamic update operators and then the path quantifiers in such formulae. Using these reductions we obtain a sound and complete axiomatization for LTC, and reduce its decision problem to that of the modal logic KD.

The Complexity of Learning Temporal Properties

Learning Branching-Time Properties in CTL and ATL via Constraint Solving

Learning Temporal Logic Properties: an Overview of Two Recent Methods

Learning Temporal Properties is NP-hard

Inferring Properties in Computation Tree Logic

Constrained LTL Specification Learning from Examples

Scalable Anytime Algorithms for Learning Fragments of Linear Temporal Logic

Learning Probabilistic Temporal Safety Properties from Examples in Relational Domains

A logic for temporal conditionals and a solution to the Sea Battle Puzzle

Adaptive Teaching of Temporal Logic Formulas to Learners with Preferences

Satisfiability Degree Computation for Linear Temporal Logic.

Standpoint Linear Temporal Logic

Learning Temporal Logic Predicates from Data with Statistical Guarantees

Linear Non-Transitive Temporal Logic, Knowledge Operations, Algorithms for Admissibility

DeepLTL: Learning to Efficiently Satisfy Complex LTL Specifications

Deep Reinforcement Learning with Temporal Logics

Information-Guided Temporal Logic Inference with Prior Knowledge

Counterfactuals Modulo Temporal Logics

Playing to Learn, or to Keep Secret: Alternating-Time Logic Meets Information Theory

Automata Linear Dynamic Logic on Finite Traces

Reinforcement learning under temporal logic constraints as a sequence modelling problem