Automated reasoning

Automated reasoning uses computers to derive conclusions or verify claims from formalized knowledge by applying logical or mathematical rules.

Formal logic: The study of formal languages and rules for valid inference. Automated reasoning represents claims in formal logics and applies their inference rules.
Formal language: A language defined by a precise alphabet and rules for forming valid expressions. Automated reasoning requires statements whose syntax and structure software can parse.
Automated theorem proving: The use of algorithms to find or verify formal proofs of mathematical statements. It is the direct application of automated inference to proving theorems.
Human reasoning: The cognitive processes by which people draw conclusions, solve problems, and assess evidence. Unlike human reasoning, automated reasoning operates on explicit formal representations and programmed rules.
Decidability: The property of a logical theory or problem having an algorithm that always gives a correct yes-or-no answer. Undecidable theories place fundamental limits on complete automated decision procedures.
Inference rule: A rule that licenses deriving a conclusion from specified premises. Reasoning systems repeatedly apply these rules to obtain new statements.
First-order logic: A logic with variables, predicates, quantifiers, and rules for reasoning about objects and relations. It expresses general claims about objects and relations that reasoning systems can process.
Model checking: An algorithmic method for checking whether a finite-state system satisfies a formal specification. It uses automated search to expose violations in hardware and software designs.
Statistical inference: Methods for estimating population properties or evaluating hypotheses from data under probabilistic models. Statistical inference quantifies uncertainty from evidence rather than deriving consequences solely from formal rules.
Computational complexity: The study of the resources required to solve computational problems as their inputs grow. Even decidable reasoning tasks may demand impractical time or memory.

Linked from 2 pages

Knowledge (artificial intelligence)Narrower topic: It is the process by which AI systems draw conclusions from represented knowledge.