Particle Swarm optimization-based Neural Network method for predicting satisfaction of recipients of internet data quota assistance from the ministry of education and culture

Abstract

The great spread of the virus, marked by the increasing number of cases and deaths, has led to calls for stay at home, work from home, and distance learning during the pandemic [1]. Therefore, to continue to support distance learning, the government is deemed necessary to ensure the availability of internet data packages for educators and students for the smooth teaching and learning process. Internet data quota assistance is distributed to students, educators, and lecturers [2]. The purpose of this study is to measure the level of user satisfaction through predictions of satisfaction to assist the government in advancing education. At Pohuwato University, the number of valid and eligible recipients of internet assistance reached 578 people, and the number of lecturers was 45. The problem is that many recipients of internet data quota assistance cannot directly convey the impression they feel when using and enjoying the internet quota provided by the government. Meanwhile, the government needs to know the level of user satisfaction to continue striving to improve and advance education. Therefore, a method is needed to help predict the satisfaction of recipients of internet data quota assistance to overcome these problems. Several studies on predicting satisfaction have been widely studied [3][4][5][6]. Still, it was found that the objects, methods, data, and parameters used by several researchers were different, resulting in different predictive values and accuracy. One method often used in prediction is Neural Network with Backpropagation algorithm. This study applies the BP algorithm and PSO feature selection. The results concluded that it successfully predicts visitor satisfaction with an accuracy value of 85.00% [7]. MethodThis study discusses the prediction of satisfaction of recipients of free quota assistance from the Ministry of Education and Culture using the Neural Network with the selection of the Particle Swarm Optimization (PSO) feature.