Constructing a decision tree from data with hierarchical class labels
Expert Systems with Applications: An International Journal
Moving towards efficient decision tree construction
Information Sciences: an International Journal
Expert Systems with Applications: An International Journal
Research on multi-valued and multi-labeled decision trees
ADMA'06 Proceedings of the Second international conference on Advanced Data Mining and Applications
UniDis: a universal discretization technique
Journal of Intelligent Information Systems
| Hi-index | 0.01 |
Most decision tree classifiers are designed to keep class histograms for single attributes, and to select a particular attribute for the next split using said histograms. In this paper, we propose a technique where, by keeping histograms on attribute pairs, we achieve (i) a significant speed-up over traditional classifiers based on single attribute splitting, and (ii) the ability of building classifiers that use linear combinations of values from non-categorical attribute pairs as split criterion. Indeed, by keeping two-dimensional histograms, CMP can often predict the best successive split, in addition to computing the current one; therefore, CMP is normally able to grow more than one level of a decision tree for each data scan.CMP's performance improvements are also due to techniques whereby non-categorical attributes are discretized without loss in classification accuracy; in fact, we introduce simple techniques, whereby classification errors caused by discretization at one step can then be corrected in the following step. In summary, CMP represents a unified algorithm that extends the functionality of existing classifiers and improves their performance.