The ability to train novice operators on a variety of tasks through automated training systems provides a cost effective and timely solution for a number of applications. The training regimen required for military personnel lends itself to automated training systems. In order to overcome some of the challenges and shortfalls of the traditional intelligent training systems, a strong normative or expert model must be developed. Such a model provides a foundation by which comparison, monitoring and feedback can take place. EMT, Inc. and California State University, Hayward propose to develop a methodology that brings together both the normative model and the student model in a approach that enables individualized training based on the trainee's individual skills. This objective will be achieved by using machine learning techniques used to develop and maintain an expert learning transition contained in the student model module via two steps. First, machine learning techniques will be used to develop and maintain an expert learning transition in the student model. Second, methods for externalizing the student model will be used to provide for student-system interaction that can be compared against the normative model
