Statistical & Financial Consulting by Stanford PhD
Home Page
MART

MART (Multiple Additive Regression Trees) is an implementation of the gradient tree boosting methods for predictive data mining (regression and classification). For a detailed description of the methodology and applications, use the references below. For a quick overview of the methodology and implementation, follow the first resource link. MART has widespread usage in Medicine, Biology, Marketing and Finance.


MART REFERENCES

Friedman, J. H. (2001), Greedy function approximation: the gradient boosting machine, Annals of Statistics.

Friedman, J. H. (1999), Stochastic gradient boosting, Technical report, Stanford University.

Friedman, J. H. & Fisher, N. (1999), Bump hunting in high dimensional data, Statistics and Computing 9, pp. 123-143.

Friedman, J. H. & Meulman J. J. (2003), Multiple additive regression trees with application in epidemiology, Statistics in Medicine, Vol. 22, Issue 9, pp. 1365-1381.

Hastie, T., Tibshirani, R., & Friedman, J. H. (2008). The elements of statistical learning: Data mining, inference, and prediction. New York: Springer.

Freund, Y. & Schapire, R. (1997), A decision-theoretic generalization of online learning and an application to boosting, Journal of Computer and System Sciences 55, pp. 119-139.

Buhlmann, P. & Hothorn, T. (2008), Boosting algorithms: Regularization, prediction and model Fitting, Statistical Science.
 


MART RESOURCES

  • MART: Overview, Tutorial, Data - a web-resource by the author of MART Jerome Friedman
  • Jerome Friedman's main page

BACK TO THE
STATISTICAL ANALYSES DIRECTORY


IMPORTANT LINKS ON THIS SITE