On the greedy heuristic for matchings

Heuristics are widely employed in practice for approximating solutions to problems of combinatorial optimization. The reduced running time of an heuristic often compensates for the difference from exact optimality in the solution obtained. This has led to theoretical interest in quantifying the efficiency of heuristics. The most common measure of solution quality is the performance ratio, the worst case of the ratio of heuristic solution value to optimal solution value. Theoretical investigations have included problems which are polynomial time solvable [7, 81 as well as problems which are NP-hard.