Comparison of Crossover and Mutation Operators to Solve Teachers Placement Problem by Using Genetic Algorithm

P H P Rosa,R A Nugroho,K Pinaryanto,A M Polina,H Sriwindono

doi:10.1088/1742-6596/1566/1/012021

P H P Rosa, R A Nugroho + Show 3 more

Open Access

https://doi.org/10.1088/1742-6596/1566/1/012021

Copy DOI

Abstract

The placement of elementary school teachers is an NP-complex problem. Teacher placement can be optimized by considering several factors that influence their performance, including the distance of teacher’s residence to school, age, and gender of the teacher. This paper discusses the solution model of the problem based on genetic algorithms by finding a chromosome formation that represents the possibility of teachers placement solution, composing a population, and finding the recommended combination of two selected mutations operators and two selected crossover operators to achieve optimal results. The selected mutation operators were Reverse Sequence Mutation (RSM) and Partial Shuffle Mutation (PSM), while the selected crossover-operators were Single Point Crossover (SPX) and Ordered Crossover (OX). The combined performance of these operators is measured based on the fitness value and running time of the program. Based on experiments, it can be concluded that the combination of OX-PSM with mutation probability 1:20 gives the lowest minimum fitness value compared to other combinations of crossover and mutation operators. The running time of the combination of OX-PSM is stable in any mutation probability, ranging from 39,5 – 41 minutes.

Highlights

One effort to improve the quality of education is through the optimal placement of teachers to support the performance of teachers in schools
To improve the optimization of the teacher placement model using genetic algorithms, this study examines the performance of the Reverse Sequence Mutation (RSM) and Partial Shuffle Mutation (PSM) mutation operators and combines them with SPX and OX crossover operators
Result and Discussion To measure the performance of RSM and PSM combined with selected crossover techniques SPX and OX, several experiments were conducted with various parameters as follows: number of population = 6 chromosomes, number of generation = 5000, number of study groups which is equal to the number of teachers = 636, whereas the probability of a mutation to the crossover varies from 1:20, 1: 40, 1:60, 1:80, 1:100, 1:200, 1:300, 1:400

Summary

Introduction

One effort to improve the quality of education is through the optimal placement of teachers to support the performance of teachers in schools. This paper discusses a model solution using genetic algorithms to solve teacher placement problems, by finding a combination of mutation and crossover operators to get good results. To improve the optimization of the teacher placement model using genetic algorithms, this study examines the performance of the RSM and PSM mutation operators and combines them with SPX and OX crossover operators. 4. Result and Discussion To measure the performance of RSM and PSM combined with selected crossover techniques SPX and OX, several experiments were conducted with various parameters as follows: number of population = 6 chromosomes, number of generation = 5000, number of study groups which is equal to the number of teachers = 636, whereas the probability of a mutation to the crossover varies from 1:20, 1: 40, 1:60, 1:80, 1:100, 1:200, 1:300, 1:400.

60 Muta8ti0on Prob1a0b0ility 200

Conclusion