Vizier: Old but fast locally weighted regression for Windows
Authors: Artur Dubrawski, Mike Duggan, Mary Soon Lee, Andrew Moore, Jeff Schneider, David Stott
Vizier was written in the mid-90's and was a combination of fast algorithms for the following operations, with a Graphical User Interface driver for Windows:
- K Nearest neighbor classification
- K Nearest neighbor regression
- Kernel regression (aka Shepherd's Interpolation)
- Kernel classification
- Locally weighted linear regression
- Locally weighted polynomial regression
- Locally weighted Bayesian polynomial regression
- Automatic determination of best model and distance metric
These days we have faster algorithms, and they run under Windows and Unix, but the functionality is not quite the same as the wild world of Vizier, so we are putting it up on the web here for free download by anyone who thinks they might find it interesting or useful.
Useful Links:
Screenshot One:
Screenshot Two:
