Measuring Student Satisfaction Based on Analysis of Physical Parameters in Smart Classroom

Several factors related to the physical environment, including temperature, environmental noise, CO2 levels, air pressure, humidity, and the lecturer's vocal delivery, significantly impact student satisfaction. Evaluating, quantifying, and enhancing these factors can significantly contribute to enhancing students' satisfaction with lecture quality. The objective of this study is to assess students' satisfaction regarding the quality of lectures through an analysis of data obtained from a variety of smart devices that record physical environmental parameters. In this study, the machine learning techniques employed encompass Decision Tree (DT), Support Vector Machine (SVM), Random Forest (RF), Adaptive Boosting (AdaBoost), K-nearest neighbors (KNN), and Stochastic Gradient Descent (SGD). Additionally, the deep learning models utilized comprise Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Gated Recurrent Unit (GRU), Bidirectional Gated Recurrent Unit (Bi-GRU), Long Short-Term Memory (LSTM), and Bidirectional Long Short-Term Memory (Bi-LSTM). The findings of this study suggest that, on the whole, deep learning models demonstrate superior performance when it comes to classifying data about this particular problem, as compared to traditional machine learning models. Notably, the CNN model emerges as the top performer among all the models assessed, achieving an impressive accuracy rate of 74.62%. it is worth mentioning that the decision tree exhibits the highest accuracy among classical machine learning models.