Title :
An iterative `flip-flop´ approximation of the most informative split in the construction of decision trees
Author :
Nádas, A. ; Nahamoo, D. ; Picheny, M.A. ; Powell, J.
Author_Institution :
IBM Thomas J. Watson Res. Center, Yorktown Heights, NY, USA
Abstract :
The authors seek a fast algorithm for finding the best question to ask (i.e., best split of predictor values) about a predictor variable when predicting membership in more than two categories. They give a fast iterative algorithm for finding a suboptimal question in the N category problem by exploiting a fast algorithm for finding the optimal question in the two-category problem. The algorithm has been used in a number of speech recognition applications
Keywords :
approximation theory; decision theory; iterative methods; speech recognition; trees (mathematics); decision trees; fast iterative algorithm; iterative flip-flop approximation; predictor variable; speech recognition; suboptimal question; two-category problem; Automatic speech recognition; Decision trees; Entropy; Flip-flops; Iterative algorithms; Mutual information; Optimized production technology; Polynomials; Probability distribution; Speech recognition;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 International Conference on
Conference_Location :
Toronto, Ont.
Print_ISBN :
0-7803-0003-3
DOI :
10.1109/ICASSP.1991.150402
