TY - GEN ID - cogprints515 UR - http://cogprints.org/515/ A1 - Wexler, Mark Y1 - 1996/// N2 - The problem of inductive learning is hard, and--despite much work--no solution is in sight, from neural networks or other AI techniques. I suggest that inductive reasoning may be grounded in sensorimotor capacity. If an artificial system to generalize in ways that we find intelligent it should be appropriately embodied. This is illustrated with a network- controlled animat that learns n-parity by representing intermediate states with its own motion. Unlike other general learning devices, such as disembodied networks, it learns from very few examples and generalizes correctly to previously unseen cases. PB - AAAI Press KW - induction KW - inductive learning KW - generalization KW - action KW - external representations KW - machine learning KW - animats KW - n-parity problem TI - Embodied induction: Learning external representations SP - 134 AV - public EP - 138 ER -