Minería de datos en R aplicada a la determinación de los usos de tecnología de los estudiantes universitarios. Caso Universidad Central del Ecuador

Abstract

Abstract: The present degree work describes the data mining process in the investigation of the uses of the internet of the students of the Central University of Ecuador (UCE), both at an academic level and for fun and entertainment, through the R Studio tool 3.6.3 and the use of the CRISP-DM methodology, which is undoubtedly one of the most used methodologies in data mining processes. The development of the project was carried out through data collected through physical surveys of UCE students, between the ages of 17 to 31 years, with the aim of applying predictive and descriptive data mining techniques, such as decision trees , artificial neural networks, k-means and principal component analysis in order to find behavior patterns and trends regarding internet uses. To verify the techniques, the area under the curve, the sum of the squared error, the elbow method and the evaluation of the proportion of the corresponding accumulated variance were used.