Optimizing learning of a locomotor task: Amplifying errors as needed | IEEE Conference Publication | IEEE Xplore