R.Rayyes presented our work at WIML and NeurIPS

R.Rayyes has presented our recent work on learning robot models efficiently at WIML and NeurIPS. Online learning methods tend to require high number of samples and data acquisition is very costly. However, it its the best choice to adapt to dynamic changes of the robot. Hence, a novel online learning scheme "Multi-Stage Goal Babbling" has been developed for learning inverse kinematics and statics simultaneously and efficiently.