New in Release 2.10
- Improved composite models
The focus of this release is composite (rules and instances) models. New algorithms have been developed to find nearest neighbors more quickly, especially for applications with many attributes, and to refine the way predicted values are estimated.
- Minor bugs fixed
For composite models, distance calculations in 2.09 are flawed if a continuous attribute has the same value for every case in the training data. In this event, the nearest neighbors found by 2.09 could be incorrect.
- Release 2.10a
This revision corrects a problem that occurs when Cubist is run with hundreds of millions of cases. It also fixes a bug that could cause rules to be listed on the output in incorrect order; this bug did not affect models or results.
Here is a summary of changes in previous releases.
For Licensees Only:
Licensees who purchased Cubist since 1 January 2018 are welcome to upgrade to Release 2.10.
Click the appropriate link(s) below to download Release 2.10. You will be asked to re-enter your licence ID before running the system. If you use our public source code for reading and interpreting Cubist models, you should also download the latest version.
- Cubist (Linux):
- Cubist (Windows 7/8/10):
- Either: source code for reading/interpreting models
|© RULEQUEST RESEARCH 2020||Last updated May 2020|