Rule-based Expert Systems

A general introduction to artificial intelligence methods of measurement signal processing is given in Article 128, Nature and Scope of AI Techniques, Volume 2. Problem solving is the process of finding a solution when the path leading to that solution is uncertain. Even though we are familiar with several problem-solving techniques, in the real world, sometimes many problems cannot be solved by a technique we are familiar with. Surprisingly, for some complicated problems, no straightforward solution technique is known at all. For these problems, heuristic solution techniques may be the only alternative. A heuristic can be simplified as a strategy that is powerful and general, but not absolutely guaranteed to provide best solutions. Heuristic methods are very problem specific. Previous experience and some general rules – often called rules of thumb – could help find good heuristics easier. Humans use heuristics a great deal in their problem solving. Of course, if the heuristic does fail, it is necessary for the problem solver to either pick another heuristic, or know that it is appropriate to give up. Choosing random solutions, adopting greedy approaches, evolving the basic heuristics for finding better heuristics are just some of the popular approaches used in heuristic problem solving (Michalewicz and Fogel, 1999). Heuristic problem solving involves finding a set of rules, or a procedure, that finds satisfactory solutions to a specific problem. A good example is finding one’s way through a maze. To make the way toward the final goal, a stepby-step movement is necessary. Very often false moves are made but in most cases we solve the problem without much difficulty. For the maze problem, a simple heuristic rule could be ‘choose the direction that seems to make progress’. Another good example is the job shop scheduling problem wherein the task is to schedule Jn independent jobs, where n = {1, 2, . . . .N} on Rm heterogeneous resources and m = {1, 2, . . . .,M}, with an objective of minimizing the completion time of all the jobs and utilizing all the resources effectively. Each job Jn has processing requirement Pj cycles and resource Rm has speed of Si cycles/unit time. Any job Jn has to be processed in resource Rm, until completion. If Cj is the completion time and the last job j finishes processing, then we define Cmax = max{Cj , j = 1, . . . , N}, the makespan and Cj , as the flow-time. The task is to find an optimal schedule that optimizes the flow-time and make-span. Some simple heuristic rules to achieve this are by scheduling the Shortest Job on the Fastest Resource (SJFR), which would minimize Cj or by