next up previous contents
Next: Judging Model Quality by Up: No Title Previous: A note on graphing

Autonomous Modeling


We have been specifying what metacode to use for each data set. A question you may have been wondering is How do I figure out what metacode to use for my data? As we saw at the beginning, the metacode choice can have a big impact on how the predictions turn out. Fortunately, Vizier automates this choice for us. This section describes how that choice is made.

Before talking about the details, we'll first see how easy it is to automatically choose a good metacode.

File -> Open -> a1.mbl
Model -> Blackbox -> Launch!

Blackbox spends 60 seconds looking for a good metacode and reports what it has found when it is done. Depending on the speed of the machine you're using, it will choose something like ``L19:MN:9.'' The metacode is automatically set to the choice it has made. You can use graph to see visually whether you think it has made a wise choice. There are several other ways to evaluate its choice which will be covered in the rest of the section.

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