Info-Fuzzy Network (IFN)

The Info-Fuzzy Network (IFN) is a software tool for automated knowledge discovery in real-world databases.  IFN is based on the information-theoretic fuzzy approach to data mining and knowledge discovery, developed by Prof. Mark Last (Ben-Gurion University of the Negev) in cooperation with Prof. Oded Maimon (Tel Aviv University) and Prof. Abraham Kandel (University of South Florida).  A complete description of this comprehensive methodology is provided in the book “Knowledge Discovery and Data Mining, the Info-Fuzzy Network (IFN) Methodology” by Oded Maimon and Mark Last (Kluwer Academic Publishers, December 2000). 

The key features of the Info-Fuzzy Network include:

·      Data Pre-processing

·      Discretization of continuous attributes

·      Feature selection (identifying the most important features)

·      Data Mining

·      Extraction of input-target associations

·      Single-target and multi-target classification and prediction

·      Online data mining

·      Post Processing

·      Rule evaluation (sorting out the most informative associations)

·      Rule fuzzification and reduction (converting the associations into a compact set of linguistic rules)

·      Data quality assurance (detecting unreliable data)

Please send your questions and comments to Prof. Mark Last (email: {mlast}@bgu.ac.il, phone:  (972)-8-6461397)

Last updated: Wednesday, August 11, 2004