Abstract
Training and education of real-world tasks in Virtual Reality (VR) has seen growing use in industry. The motion-tracking data that is intrinsic to immersive VR applications is rich and can be used to improve learning beyond standard training interfaces. In this paper, we present machine learning (ML) classifiers that predict outcomes from a VR training application. Our approach makes use of the data from the tracked head-mounted display (HMD) and handheld controllers during VR training to predict whether a user will exhibit high or low knowledge acquisition, knowledge retention, and performance retention. We evaluated six different sets of input features and found varying degrees of accuracy depending on the predicted outcome. By visualizing the tracking data, we determined that users with higher acquisition and retention outcomes made movements with more certainty and with greater velocities than users with lower outcomes. Our results demonstrate that it is feasible to develop VR training applications that dynamically adapt to a user by using commonly available tracking data to predict learning and retention outcomes.
Highlights
Virtual Reality (VR) has been used extensively for education and training [1]
It is clear from the discrepancies in high-performance accuracy between the training (High Acc = 1.000) and testing (High Acc = 0.000) datasets and the Matthews correlation coefficient (MCC) scores for the testing data (MCC = −0.135) that the first and third sets of input features, both positionbased, suffered from overfitting the high-performance training data
The velocity-based fifth set yielded the best test MCC of 0.674 and a better overall accuracy of 0.833 than the position-based second set, which yielded a test MCC of 0.400 and an overall accuracy of 0.783. This position-based second set was more conservative in identifying low-performance participants in both the training (Low Acc = 0.725) and testing (Low Acc = 0.900) datasets than the velocitybased fifth set, which was more accurate for both training (Low Acc = 0.825) and testing (Low Acc = 1.000)
Summary
Virtual Reality (VR) has been used extensively for education and training [1]. VR practitioners have developed educational VR environments for knowledge acquisition, such as learning about geometry [2], World War I [3], and how to inspect a haul truck [4].VR has been used for training psychomotor skills, such as how to visually scan for threats [5], use measurement tools [6], and put on personal protective equipment [7]. The field of Intelligent Tutoring Systems (ITSs) is closely related to VR training and education. ITSs incorporate Artificial Intelligence (AI) techniques to allow for tutoring systems that know “what they teach, who they teach, and how to teach it” through user models [11]. ITSs can use these user models to provide adaptive support for learners because they provide a representation of the learner in terms of relevant traits like learning behaviors and metacognitive ability [12]. This adaptive support can be implemented via proactive guidance (scaffolding) or reactive guidance (feedback) [13]. Most ITSs that create adaptive learning environments employ feedback [13], which is usually implemented with static rules for specific types of detected errors [14]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have