(of an approach to problem solving, learning, or discovery) That employs a practical method not guaranteed to be optimal or perfect; either not following or derived from any theory, or based on anadvisedlyoversimplified one.[from 1821]
2015, Emiliano Ippoliti, Thomas Nickles,Heuristic Reasoning[1]:
"Theheuristic appraisal is the 'identification and evaluation of hints and clues that can provide direction to inquiry in the sometimes large gap between the extremes of complete knowledge and complete ignorance'”.
(computing, of a method or algorithm) That provides a useful, but notoptimal, solution to a problem. Such algorithms are typically employed either because the only known algorithms that provide optimal solutions use too much time or resources, or else because there is no known algorithm that provides an optimal solution.
2002, Te Chiang Hu, Man-tak Shing,Combinatorial Algorithms[2]:
If aheuristic algorithm works for most of the input data or its maximum percentage error is tolerable, we may prefer theheuristic algorithm to an optimum algorithm that requires a long time.
(of an argument) That reasons from the value of a method or principle that has been shown by experimental investigation to be a useful aid in learning, discovery and problem-solving.
Heuristic (computer science) on Wikipedia.Wikipedia