Next: Practical details
Up: Experimental Evaluation
Previous: Experimental Evaluation
Contents
There are many questions that can be asked about TClass'
performance such as:
- How accurate is TClass on classification tasks?
- How comprehensible are the results and definitions generated by
TClass?
- How much data does TClass need to learn concepts (i.e.,
data efficiency)?
- How does TClass compare to other learning methods in
terms of performance, comprehensibility, execution time and data
efficiency?
- How do the the settings of TClass' parameters (eg. number
of random subsets to use, whether we use smoothing or not, etc.)
affect its performance?
- How do different TClass back end learners perform?
Subsections
Mohammed Waleed Kadous
2002-12-10