Robot learning by nonparametric regression

We present an approach to robot learning based on a nonparametric regression technique, locally weighted regression. The model of the task to be performed is represented by infinitely many local linear models, i. e. , the (hyper-) tangent planes at every query point. Such a model, however, is only generated when a query performed and is not retained. The architectural parameters of our approach, such as distance metrics, are also a function of the current query point instead of being global. Statistical tests are presented for when a local model is good enough such that it can be reliably used to build a local controller. These statistical measures also direct the exploration of the robot. We explicitly deal with the case where prediction accuracy requirements exist during exploration. By gradually shifting a center of exploration and controlling the speed of the shift with local prediction accuracy, a goal-directed exploration of state space takes place along the fringes of the current data support until the task goal is achieved. We illustrate this approach by describing how it has been used to enable a robot to learn a challenging juggling task. >

Authors

• Stefan Schaal
• Christopher G. Atkeson

Keywords

• Orbital robotics
• Cognitive robotics
• Learning
• Testing
• State-space methods
• Artificial intelligence
• Laboratories
• Intelligent robots
• Piecewise linear techniques
• Regression analysis
• Nonparametric Regression
• Robot Learning
• Linear Model
• State Space
• Statistical Measures
• Learning System
• Task Model
• Distance Metrics
• Starting 2
• Tangent Plane
• Piecewise Linear Model
• Query Point
• Set Of Linear Models
• Local Linear
• Training Data
• High-dimensional
• Dimensional Space
• Source Code
• Optimal Control
• Inverse Model
• Prediction Intervals
• Forward Model
• Van Zyl
• Instance-based Learning
• Nominal Position
• Partition Tree
• Input Space
• Input Dimension
• Cross-validation Error