2017 Kuefler Geometric Concept Acquisition
<p>January 2, 2020, 1:45 PM</p>
<p> </p>
<p>Alex Kuefler and Mykel Kochenderfer and Jay.</p>
<p> </p>
<p>Explaining how intelligent systems come to <strong>embody knowledge of deductive concepts through inductive learning </strong>is a fundamental challenge of both cognitive science and artificial intelligence.</p>
<p> </p>
<p>We address this challenge by exploring how a <strong>deep reinforcement learning agent, occupying a setting similar to those encountered by early-stage mathematical concept learners, comes to represent ideas such as rotation and translation.</strong> We first train a Dueling Deep Q-Network on a <strong>shape sorting task requiring implicit knowledge of geometric properties,</strong> then we query this network with classification and preference selection tasks. We demonstrate that <strong>scalar reinforcement provides sufficient signal to learn representations of shape categories.</strong> After training, the model shows a preference for more symmetric shapes, which it can sort more quickly than less symmetric shapes, supporting the view symmetry preferences may be acquired from goal-directed experience.</p>