Abstract
FID3.2 is a public software tool for supervised classification, implementing the popular and efficient recursive partitioning technique of decision trees, while utilizing fuzzy representation and approximate reasoning for dealing with noise and language uncertainty. It is loosely based on the ID3 decision tree algorithm. It handles nominal, continuous, and linguistic attributes and classes. It can also operate with noisy and unknown features, both in training and in testing data. For continuous attributes that are not pre-partitioned, the system generates fuzzy partitioning using either a top-down or a bottomup method. Finally, FID uses a number of inferences from two classes: set-based and exemplar-based. Recently, we have implemented a GUI interface to work with FID3.2. The interface provides graphical visualization of data, partitions, and the tree, and it also allows limited graphical manipulations. This paper describes the GUI interface
Original language | American English |
---|---|
Journal | North American Fuzzy Information Processing Society |
DOIs | |
State | Published - Jun 1 2007 |
Disciplines
- Theory and Algorithms
- Computer Sciences
- Artificial Intelligence and Robotics