Análise comparativa de algoritmos de aprendizagem de máquina para classificação de dificuldade de cursos do tipo mooc

Abstract

The current work involves the development of a Data Analysis and Machine Learning application to classify Massive Open Online Courses (MOOCs), which are online, open, and massive courses, into different levels of difficulty. Algorithms such as Random Forest Classifier, Logistic Regression, and K-neighbors classifier are used to determine the best classification approach. The objective is to find the optimal algorithm settings, considering different metrics. Proper classification of the courses is crucial to avoid frustrations and loss of trust from the participants. The application was designed to integrate into an online course platform in the future, with Coursera being chosen as the international reference. Using Coursera’s dataset, the application provides more correct classification of the courses’ difficulty level based on the required skills, course name, and responsible institution, assisting course creators in proper classification and ensuring course suitability for participants.