Robot Shaping: An Experiment in Behavior Engineering

摘要: From the Publisher: foreword by Lashon Booker "[This] book gives a clear and comprehensive exposition of [the authors] extensive experience in integrating reinforcement learning autonomous robotics. Their continuing contribution is to development distinct engineering discipline (`Behavior engineering') through which such robots can be created. I am excited because their efforts combine some best theoretical ideas with strong eye for practical - what will actually work." -- Stewart W. Wilson, The Rowland Institute Science To program an robot act reliably dynamic environment complex task. dynamics are unpredictable, robots' sensors provide noisy input. A robot, one that acquire knowledge interaction its then adapt behavior, greatly simplifies designer's work. need not given all details environment, actuators finely tuned. Robot Shaping about designing building robots. term "shaping" comes from experimental psychology, where it describes incremental training animals. authors propose new discipline, "behavior engineering," methodologies tools creating techniques based on classifier systems, architecture originated John Holland, they have added several ideas, as "mutespec," system "energy," population size. In present Behavior Analysis Training (BAT) example behavior methodology.