Population Based Techniques for Solving the Student Project Allocation Problem

Abstract

The student project allocation problem is a well-known constraint satisfaction problem that involves assigning students to projects or supervisors based on a number of criteria. This study investigates the use of population-based strategies inspired from physical phenomena (gravitational search algorithm), evolutionary strategies (genetic algorithm), and swarm intelligence (ant colony optimization) to solve the Student Project Allocation problem for a case study from a real university. A population of solutions to the Student Project Allocation problem is represented as lists of integers, and the individuals in the population share information through population-based heuristics to find more optimal solutions. All three techniques produced satisfactory results and the adapted gravitational search algorithm for discrete variables will be useful for other constraint satisfaction problems. However, the ant colony optimization algorithm outperformed the genetic and gravitational search algorithms for finding optimal solutions to the student project allocation problem in this study.

Meta-Heuristic Solutions to a Student Grouping Optimization Problem faced in Higher Education Institutions

“…Heuristic algorithms have successfully been used to solve NP-hard problems such as the Student Project Allocation Problem [7,8,9] and the Nurse Rostering Problem [10,11].…”

“…Magalhães-Mendes [22] compared the efficiency of different crossover approaches for a job scheduling problem, with experiments showing the single point crossover to produce the best average performance. However, in a study [7] that solved the student project allocation problem, the type of crossover operator did not have any significant influence on the performance of the genetic algorithm, emphasizing that no method is guaranteed to outperform others in all problems.…”

“…The techniques [5] reviewed can be transferred to solving other grouping problems faced in Higher Institutions that such as the Student Problem Allocation Problem have been shown to be NP-hard [6]. For example, Manlove et al [6] has solved a Student Project Allocation Problem with Integer Programming and Kenekayoro et al [7] found optimal solutions with Ant Colony Optimization, Genetic Algorithm and the Gravitational Search Algorithm.…”

“…A number of variations to the ACO exist and have been used to solve meta-heuristic problems such as the University Course Timetabling Problem [19,4], Student Project Allocation Problem [7], Nurse Rostering Problem [20]. However, the following concepts as described in [7] are the basis of all ACO algorithms.…”

A Meta-Heuristic Search Algorithm based on Infrasonic Mating Displays in Peafowls

“…It is necessary to find optimal combinations for the efficient management of scarce resources, which is why optimization methods are useful and have been applied in a number of fields among which include Manufacturing [2,3], Engineering [4,5] and Education [6,7].…”