Education 4.0: Explainable Machine Learning for Classification of Student Adaptability
Raj Gaurang Tiwari, Anuj Kumar Jain, Vinay Kukreja, Neha Ujjwal
Abstract
Because of the Covid 19 outbreak, which forced the closure of schools and institutions in India after March 2020, online education gained even greater traction in the country. The flexibility of online education allows students to rewatch recorded lectures as many times as they need to fully comprehend the topic. However, there are challenges and ethical concerns associated with using this technology in the classroom. Future possibilities, benefits, and drawbacks of using AI in the classroom are yet to be explored. Despite the great popularity and efficiency of online education, some studies have demonstrated that it may be detrimental to students. This research investigates students’ attitudes on the use of technology in the classroom. We use the idea of Explainable Machine Learning (XML), in which the outcomes of ML calculations are explicable to people. On the other hand, the “black box” approach holds that not even the creators of an AI can explain how it arrived at a certain choice. Several machine learning methods were used to forecast the extent to which students will embrace Industry 4.0 features. The most successful method was shown to be Neural Network (NN) which achieved an impressive 93% accuracy in classification. By detailing the inner workings of models to give some level of explainability, we can fully grasp the promise of this algorithm.