Heuristic Search Strategies-Artificial intelligence
In general speaking, a heuristic search is one which utilizes a rule of thumb to improve an agent's performance in solving problems by search. A heuristic search is not to be confused with a heuristic measure .As discussed below, if you may identify a heuristic measure, then it opens up a range of generic heuristic searches which you can try to improve your agent's performance. However, it is importance remembering that any law of thumb, for instance, selecting the order of operators is a heuristic ,when applied in a simple breadth first search.
In terms of our plan analogy, a heuristic search select where to put a (state, operator) pair on the plan when it is proposed as a move in the state space. This option could be complexes and based on many factors. In the terms of the graph analogy, a heuristic search selects that node to expand at any point in the search. By definition, a heuristic search is not guaranteed to get better performance for a particular problem or set of problems, but they all are implemented in the hope of improving the speed of which a solution is found or /and the quality of the solution searched. In fact, we may be able to search optimal solutions, which are as good as possible with reference to some measure.
Optimality
The path worth of a solution is calculated as the addition of the costs of the actions which led to that solution. This is only one example of a measure of value on the solution of a search problem, and there are many others. These measures may /may not be connected to the heuristic functions which estimate the likelihood of a specific state being in the path to a solution. We say that given a measure of value on the possible solutions to a search problem - 1 specific solution is best possible if it gain higher than all the others with respect to this measure (or costs less, in the case of path cost). I.e. in the maze example given in section 3.2, there are several paths from the start to the finish of the maze, but just one which crosses the fewest squares. This is the good solution in terms of the distance travelled.
Optimality may be guaranteed through a particular option of search strategy (for instance the uniform path cost search described below). On the other hand, an agent may select to prove that a solution is optimal by appealing to some mathematical argument. As a final resort, if optimality is required, then an agent might exhaust a full search strategy to find all solutions, then choose the one scoring the highest (alternatively costing the lowest).