Titarl | Experiment manager

Introduction

Titarl can used in two different ways. The first way is to use Titarl with the command line (e.g. --learn, --apply, --display, --eval). This solution is simple but becomes fastidious, when changing and comparing Titarl parameters, when applying cross-validation, or to stop and resume runs. The second solution is to use the Titarl Experiment Manager that will calls Titarl command line itself. The second solution is preferable for more complex analysis.

Using the Experiment Manager offers the following adventages:

• Compact high level configuration file.
• Easy cross-validation evaluation of the rules.
• Caching of the intermediate results to help resuming or modifing an analysis.
• Easy comparison of different algorithm results (including Titarl) and different parameter values.
• Clear organisation of the results (rules, predictions, evaluation, etc.).

Note that when using the Experiment Manager, you can look in the log files to see which Titarl command were executed.