The Application of Clustering Analysis in the Assessment of Eye Movements During Flight Training Intervention

Abstract

In this research, the application of K-means clustering analysis on pilots’ gaze points was studied in order to develop a quick evaluation approach for gazed-based training interventions. Detailed analysis of eye movements’ data takes a considerable amount of time which makes it difficult to use in some stages of training intervention. As a result, a more efficient indicator of trainees’ performance is needed to allow trainers the ability to provide timely feedback on trainees’ performance. In this research, the potential of using clustering analysis on pilots’ gaze points was assessed in order to classify bad and good gaze behaviors. One expert pilot and six novice pilots participated in the experiment. Pilots were equally divided into experimental and control groups.