I have developed a greedy graph search algorithm and made some optimizations. However, its performance varies depending on the parameters I use. I am keen to create a module that can compute the optimal parameter settings to enhance the efficiency of my algorithm. Are there any techniques available to precisely analyze the amount of searching involved in greedy graph searching (approximately exact)? I prefer not to rely on "time complexity" analysis as it may not provide sufficient precision for my specific problem.
I want to build a module that can compute the optimal parameter settings to enhance the efficiency of my algorithm.