Success history intelligent optimizer

Hussam N. Fakhouri*, Faten Hamad, Abedalsalam Alawamrah

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Citations (Scopus)


Swarm intelligence (SI) is part of artificial intelligence that is based on the collective behavior of particles in a self-organized and decentralized intelligent systems which is inspired from either the human or animal social behavior. SI has a great involvement in solving optimization problems. As optimization-based systems are complicated, so powerful decentralized algorithms supported by SI are used to solve such problems. Intelligent algorithms of SI can solve the challenging issues of optimization due to their different properties; thus, there is an increased demand to enhance the performance and propose new and novel SI algorithms. This paper presents a novel stochastic swarm intelligence algorithm called success history intelligent optimizer (SHIO). It offers a solution for single-objective optimization problems by proposing a new exploration and exploitation movement strategy based on the three best solutions found in the search space to create a new movement vector, where each best solution is stored in the memory and subtracted from the average of the three best solutions found so far during the optimization process. The proposed SHIO ensures the efficiency of search space exploration and use. In order to confirm SHIO performance, several performance measurements (search history, trajectory and convergence curves) have been tested and SHIO was used to solve (23) single-objective optimization benchmarking functions. These functions have been classified to unimodal, multimodal and multimodal fixed. Various metrics such as mean, standard deviation, minimum and maximum have been utilized, and quantitative findings have been recorded. Further, trajectory and search history of the qualitative result were visualized. The results of test functions and performance metrics demonstrate that the proposed algorithm can explore various search area locations, make use of potential search space locations while optimizing, avoid local optimism and converge to the global best efficiently. SHIO delivers highly competitive and superior results in the evaluated unimodal and multimodal benchmarks over the compared algorithms. Note that SHIO algorithm source code is available on and open for public use.

Original languageEnglish
Article number5
Pages (from-to)6461-6502
Number of pages42
JournalJournal of Supercomputing
Issue number5
Publication statusPublished - Oct 19 2021
Externally publishedYes


  • Decentralized intelligent systems
  • Metaheuristics
  • Optimization
  • Single-objective optimization
  • Swarm intelligence

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Software
  • Information Systems
  • Hardware and Architecture

Cite this