DMS Home

DM Methodology

Modelling techniques

Modelling techniques come from fields of machine learning, signal processing, evolutionary computing and statistics. Techniques are one thing that makes data mining so specific. Modelling techniques in data mining have also one important characteristic - they au tomaticaly generate new propositions (models), about relations among important variables in the data, which is an added value to traditional confirmatory statistical analyses. Due to the essential impact that modelling techniques have on DM process, in the table bellow you will find links to parts of the tutorial devoted to more thorough presentation of basic features of data mining modelling techniques and their application.


Common structure of DM modelling techniques Model representation
Estimation criteria
Search method
Description of DM modelling techniques Decision trees
Rule Induction Methods
Association Rules
Clustering Methods
Neural Networks
ILLM Rule Induction system
Evaluation of generated models Detailed description of evaluation measures and techniques for evaluating classifier models.




© 2001 LIS - Rudjer Boskovic Institute
Last modified: March 28 2017 05:23:30.