Finding an Optimal Team
Michał Okulewicz · Annals of Computer Science and Information Systems · 2016
This article proposes a metaheuristic optimization/social simulation approach to find the optimal team for a given type of the project.The quality of the team is assessed in a black-box optimization environment, where the optimized function acts as a metaphor of the project to be completed within the certain time limit (number of fitness function evaluations) and each fitness function evaluation is considered to be a metaphor of a unit task.The employees in a team are modeled according to the Belbin's Team Roles and the Particle Swarm Optimization (PSO) is used as a teamwork framework algorithm, while Evolutionary Algorithm (EA) as an algorithm for controlling the set of Team Roles for team members and leaders.This approach has been tested in a scenario of a simulated self-organizing team, where each employee decides about his own actions.The results from the performed simulation suggest, that such teams perform best if their leader is one of the actual work-oriented roles.Additionally, some projects required significantly different set of roles than the average team, resulting in improvement of the specialized team's performance over that of the average team.