### Abstract

Combinatorial optimization focuses on arriving at a globally optimal solution given constraints, incomplete information and limited computational resources. The combination of possible solutions are rather vast and often overwhelms the limited computational power. Smart algorithms have been developed to address this issue. Each offers a more efficient way of traversing the search landscapes. Critics have called for a realignment in the bio-inspired metaheuristics field. We propose an algorithm that simplifies the search operation to only randomized population initialization following the Randomized Incremental Construction Technique, which essentially compartmentalizes optimization into smaller sub-units. This relieves the need of complex operators normally imposed on the current metaheuristics pool. The algorithm is more generic and adaptable to any optimization problems. Benchmarking is conducted using the traveling salesman problem. The results are comparable with the results of advanced metaheuristic algorithms. Hence, suggesting that arbitrary exploration is practicable as an operator to solve optimization problems.

Language | English |
---|---|

Pages | 11-20 |

Number of pages | 10 |

Journal | Evolutionary Intelligence |

Volume | 12 |

Issue number | 1 |

Early online date | 18 Sep 2018 |

DOIs | |

Publication status | Published - 1 Mar 2019 |

### Fingerprint

### Keywords

- Bio-inspired algorithms
- Combinatorial optimization
- Randomized incremental construction
- Traveling salesman problem

### Cite this

}

**The Anglerfish algorithm: A derivation of randomized incremental construction technique for solving the traveling salesman problem.** / Pook, Mei F.; Ramlan, E.

Research output: Contribution to journal › Article

TY - JOUR

T1 - The Anglerfish algorithm: A derivation of randomized incremental construction technique for solving the traveling salesman problem

AU - Pook, Mei F.

AU - Ramlan, E.

N1 - Started Ulster 2 May 2019 (Source: CoreHR). Was previously employed at University Malaya (see other file Ramlan 76781929 Author Affiliation).

PY - 2019/3/1

Y1 - 2019/3/1

N2 - Combinatorial optimization focuses on arriving at a globally optimal solution given constraints, incomplete information and limited computational resources. The combination of possible solutions are rather vast and often overwhelms the limited computational power. Smart algorithms have been developed to address this issue. Each offers a more efficient way of traversing the search landscapes. Critics have called for a realignment in the bio-inspired metaheuristics field. We propose an algorithm that simplifies the search operation to only randomized population initialization following the Randomized Incremental Construction Technique, which essentially compartmentalizes optimization into smaller sub-units. This relieves the need of complex operators normally imposed on the current metaheuristics pool. The algorithm is more generic and adaptable to any optimization problems. Benchmarking is conducted using the traveling salesman problem. The results are comparable with the results of advanced metaheuristic algorithms. Hence, suggesting that arbitrary exploration is practicable as an operator to solve optimization problems.

AB - Combinatorial optimization focuses on arriving at a globally optimal solution given constraints, incomplete information and limited computational resources. The combination of possible solutions are rather vast and often overwhelms the limited computational power. Smart algorithms have been developed to address this issue. Each offers a more efficient way of traversing the search landscapes. Critics have called for a realignment in the bio-inspired metaheuristics field. We propose an algorithm that simplifies the search operation to only randomized population initialization following the Randomized Incremental Construction Technique, which essentially compartmentalizes optimization into smaller sub-units. This relieves the need of complex operators normally imposed on the current metaheuristics pool. The algorithm is more generic and adaptable to any optimization problems. Benchmarking is conducted using the traveling salesman problem. The results are comparable with the results of advanced metaheuristic algorithms. Hence, suggesting that arbitrary exploration is practicable as an operator to solve optimization problems.

KW - Bio-inspired algorithms

KW - Combinatorial optimization

KW - Randomized incremental construction

KW - Traveling salesman problem

UR - http://www.scopus.com/inward/record.url?scp=85053682890&partnerID=8YFLogxK

U2 - 10.1007/s12065-018-0169-x

DO - 10.1007/s12065-018-0169-x

M3 - Article

VL - 12

SP - 11

EP - 20

JO - Evolutionary Intelligence

T2 - Evolutionary Intelligence

JF - Evolutionary Intelligence

SN - 1864-5909

IS - 1

ER -