Resolving team selection in agile development using NSGA-II algorithm

Agile development methodology leads the software organization in development of software products. Development of software products depends on team members who has the ability to compete. Selection of team members for development, involves a tedious and complex process. In the current scenario, selection of individuals to form a team, based on various parameters, is a time consuming process. The time consumption can be limited, by selecting teams based on average team cost and team productivity to form a new team is discussed in this work. Team selection problem can be solved using NSGA-II algorithm, which is multi-objective meta-heuristic. This algorithm is chosen because this approach allows an easy specification of rules to choose the best team. The best team is selected by applying mutation, cross over operations on the random initial population. New offspring produced is combined with existing population to obtain a new population. New population are ranked by applying fast non-dominated sorting algorithm. Based on the rank, population are segregated into pareto frontiers. Crowding distance is applied to calculate the distance among the frontiers. The highest crowding distance are chosen to form a team.