On the Computational Economics of Reinforcement Learning

Following terminology used in adaptive control , we distinguish between indirect learning methods, which learn explicit models of the dynamic structure of the system to be controlled , and direct learning methods, which do not. We compare an existing indirect method, which uses a conventional dynamic programming algorithm, with a closely related direct reinforcement learning method by applying both methods to an innnite horizon Markov decision problem with unknown state-transition probabilities. The simulations show that although the direct method requires much less space and dramatically less computation per control action, its learning ability in this task is superior to, or compares favorably with, that of the more complex indirect method. Although these results do not address how the methods' performances compare as problems become more diicult, they suggest that given a xed amount of computational power available per control action, it may be better to use a direct reinforcement learning method augmented with indirect techniques than to devote all available resources to a computation-ally costly indirect method. Comprehensive answers to the questions raised by this study depend on many factors making up the economic context of the computation.