Heuristics from Nature for Hard Combinatorial Optimization Problems

Abstract

In this paper we try to describe the main characters of Heuristics ‘derived’ from Nature, a border area between Operations Research and Artificial Intelligence, with applications to graph optimization problems. These algorithms take inspiration from physics, biology, social sciences, and use a certain amount of repeated trials, given by one or more ‘agents’ operating with a mechanism of competition‐cooperation. Two introductory sections, devoted respectively to a presentation of some general concepts and to a tentative classification of Heuristics from Nature open the work. The paper is then composed of six review sections: each of them concerns a heuristic and its application to an NP‐hard combinatorial optimization problem. We consider the following topics: genetic algorithms with timetable problems, simulated annealing with dial‐a‐ride problems, sampling and clustering with communication spanning tree problems, tabu search with job‐shop‐scheduling problems, neural nets with location problems, ant system with travelling salesman and quadratic assignment problems.

Parametric Study for an Ant Algorithm Applied to Water Distribution System Optimization

“…The parameters τ 0 and Q appear only together as a ratio in (28), and as (28) represents Table 4 Sensitivity analysis summary for pheromone persistence factor ρ with α = 1.0, β = 0.5, m = 100, I max = 1,000 for the NYTP and HP and 10,000 for the NYTP2 and τ 0 = 100, 50, and 200 for the NYTP, HP, and NYTP2, respectively. Table 5 Table 5 …”

A Hybrid Social Network-based Collaborative Filtering Method for Personalized Manufacturing Service Recommendation

“…For instance, Sobecki [24] proposed five swarm intelligence algorithms in the field of student course recommendation, including ant colony optimization (ACO) [5] and PSO. Specifically, PSO was employed to find the set of the optimal neighborhood of students for further grade prediction.…”

An Ant Colony Approach to the Orienteering Problem

“…The Ant System (AS) was first introduced by Dorigo and his colleagues [18,19,20,23]. Since then, ACO algorithms have been applied to different problems such as the traveling salesman problem (TSP) [4,6,9,10,11,21,22,23,25,30,46], the quadratic assignment problem (QAP) [40,52], the generalized assignment problem [48], the vehicle routing problem [5,7,16,17,27], telecommunication networks [15], graph coloring [13,34,35], scheduling [1,12,14,32,42,43], the shortest supersequence problem [44,45], the Hamiltonian graph problem [55], the multiple knapsack problem [37], the sequential ordering problem [26], the redundancy allocation problem [39], water distribution network design [41], the constraint satisfaction problem [51], and continuous function problems [2,3,57]. Local search plays an important role in improving the solution quality of ACO algorithms.…”