It is very interesting to provide a framework for abstracting human skills so as to facilitate analysis of human behaviors, to allow machines to learn from human partners in the human-machine cooperation, and to transfer skill from human to human through learning human machine interfaces.
The research issues that we are addressing are:
(1) Efficiently model human control strategy;
(2) Validate the computed model;
(3) Evaluate the quality of the model;
(4) Select the input space in order to generate reliable model;
(5) Transfer human control strategy effectively.
This work has potential application in a number of different areas. A better understanding and modeling of Human Control Strategies(HCS) can lead to automatic safety devices for cars and other human-operated machines.
By means of HCS models, these devices can alert the human operator once a critical situation occurs or when the human’s performance deteriorates, for example a driver begins to fall asleep.
In the video-game industry as well as the emerging virtual-reality market, the research can provide customers with more realistic and exciting “smart” games, which incorporate techniques
developed herein. Rather than to treat each player equally the same, video and virtual-reality games can “learn” the skill level of individual users, and consequently “adjust” the game difficulty level accordingly.
| 
