Boosting   "This material is based upon work supported by the National Science Foundation under Grant No. 0102636." "Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation."

Background ----- See R. Schapire's homepage as well as an online resource on boosting www.boosting.org.

Technical Reports
(A set of papers, with mutual overlaps, on some theoretical aspects of boosting: on the assumption of weak hypotheses; behavior of generalization error in the large time limit and during the process of boosting with comparison to the optimal Bayes error; performance in the noiseless and noisy situations; overfitting and regularization; analogy of regression and classification boosting algorithms.)
(An overview of these results is contained in this brief survey.)
 

  • On Weak Base Hypotheses and Their Implications for Boosting Regression and Classification.
    Technical Report 00-01, Department of Statistics, Northwestern University. PostScript
    (Submitted 11/2/1999. Revised 10/31/2000.) (Appeared in The Annals of Statistics, 2002, 51-73.)
    (Former title: `Large Time Behavior of Boosting Algorithms for Regression and Classification'.
    Technical Report 99-01, Department of Statistics, Northwestern University. PostScript)
    •  
  • On Weak Base Learners for Boosting Regression and Classification.
    Technical Report 00-02, Department of Statistics, Northwestern University. PostScript
    (Submitted 1/16/1999 to MCS 2000 and a condensed version appeared in Proceedings: The First International Workshop on Multiple Classifier Systems, June 2000, 87-96. Lecture Notes in Computer Science 1857, Springer.)
    •  
  • Does Boosting Overfit: Views from an Exact Solution.
    Technical Report 00-03, Department of Statistics, Northwestern University. PostScript
    (Submitted 9/22/2000.)
    •  
  • Is Regularization Unnecessary for Boosting?
    Technical Report 00-04, Department of Statistics, Northwestern University. PostScript
    (Submitted 11/2/2000 to AISTATS 2001. To appear in Proceedings: The Eighth International Workshop on Artificial Intelligence and Statistics, January 2001, Morgan Kaufmann.)
    •  
  • Process Consistency for AdaBoost.
    Technical Report 00-05, Department of Statistics, Northwestern University. PostScript
    (Submitted 11/30/2000.) (To appear in The Annals of Statistics.) (With discussions and a rejoinder. PostScript)
    •  
  • Some Theoretical Aspects of Boosting in the Presence of Noisy Data.
    Technical Report 01-01, Department of Statistics, Northwestern University. PostScript
    (A brief survey. Submitted 1/16/2001 to ICML 2001. To appear in Proceedings: The Eighteenth International Conference on Machine Learning, June 2001, Morgan Kaufmann.)

    •  

    (back to the referring page)

    Return to main NORTHWESTERN Statistics page

    (last updated on 3/20/2001.)