Eyes, Brain, Emotion: AI-Powered Learning Prediction

Abstract

This study aimed to predict learning performance using machine learning based on learners’ brain waves, eye fixations, and emotional states while studying a multimedia material designed with positive emotional elements. Fifty-nine university students (27 women, 32 men; M = 20.2) participated. During learning, eye movements were recorded with an eye-tracking device, and brain signals were captured via EEG. Learners’ emotional states were measured using the Positive and Negative Affect Schedule, and learning performance was assessed through a retention test. Multiple Linear Regression (MLR) and Artificial Neural Network (ANN) methods were employed to predict learning outcomes. MLR results revealed that positive emotion and fixation were significant predictors of learning performance. Using all input variables, the ANN model achieved high prediction accuracy, suggesting that integrating neurophysiological and affective data can effectively model learning performance in emotionally designed multimedia environments.