Greedy Algorithms

Abstract

Greedy algorithms seek to find optimal solutions by making locally optimal steps. They form a class of usually simple (simple from the point of view of both description and time complexity) algorithms which often return optimal solutions or have good enough accuracy. This article presents representative greedy algorithms returning optimal solutions, such as Dijkstra's, Prim's, Huffman's, and Kruskal's algorithms. In addition, matroids, greedy algorithms for NP-hard problems, the set cover problem as well as randomized greedy algorithms and greedy approaches for continuous models are also discussed.