A Comparison of Machine Learning Algorithms for Internet Cost in Satun Education Sandbox

Satun Province is a pilot in driving innovation in education in the area. Also known as Satun Education Sandbox. Sixteen schools participate in the project located in Satun Province, the south of Thailand. During the Covid-19 situation, many students study at home. The emergence of online learning increases the internet cost per family. Parents and teachers are aware of this cost. Moreover, the Satun Education Sandbox committee need to draw the strategies from the existed data. In this research, the data was collected from a survey of 2,594 people including teachers and students about the media that used in the teaching and learning during Covid-19 and the internet cost per family. This paper will compare machine learning algorithm, including LR, LDA, k-NN, CART, SVM, Naïve Bayes, SVM, RF, and MLP algorithms, to foresight the cost. After cleaning and some more precise configuration of the results, the details of the data set are described in detail. The model can then predict the expected cost of household internet use at home. It will be helpful for the Ministry of Education plan for further assistance to students in online learning.