Publications

Previous | item21 | Next

Here is a summary of publications on Pruning methods, as discussed in datamine-l list.

For CART the definitive book is Leo Breiman et al., Classification and Regression Trees (CART), Wadsworth & Brooks, 1984. For C4.5 see J. Ross Quinlan, C4.5: Programs for Machine Learning, Morgan Kaufmann, 1993.

Summaries for at least some algorithms can be found in D. Michie, D.J. Spiegelhalter and C.C. Taylor (eds.), Machine Learning, Neural and Statistical Classification Ellis Horwood, 1994. Another good overview book is D. J. Hand, Construction and Assessment of Classification Rules, Wiley, 1997. Can be downloaded in postscript or PDF format from http://www.amsta.leeds.ac.uk/~charles/statlog/.

See also http://www.amstat.org/publications/jcgs/abstracts98/leblanc.html Monotone Shrinkage of Trees Michael LeBlanc and Robert Tibshirani

Abstract: We investigate a new method for regression trees which obtains estimates and predictions subject to constraints on the coefficients representing the effects of splits in the tree. The procedure leads to both shrinking of the node estimates and pruning of branches in the tree and for some problems gives better predictions than cost- complexity pruning used in the classification and regression tree (CART) algorithm. The new method is based on the least absolute shrinkage and selection operator (LASSO) method developed by Tibshirani.

Previous | item21 | Next