next up previous contents
Next: Choosing Parameters to Achieve Up: No Title Previous: Autonomous model selection pros

Decisions with Locally Weighted Models


In this tutorial we have seen how locally weighted learning works, and how to automatically select good locally weighted models. We have also seen how to make predictions and get confidence intervals from the models, and how to graph the results. As we have stated before, though, making predictions is not what learning is about. It is about making decisions. In this section we will demonstrate how LWL and Vizier are used to make decisions. For more information on using LWL for control see [5].

Jeff Schneider
Fri Feb 7 18:00:08 EST 1997