Logistic regression, AdaBoost and Bregman distances

Machine Learning

Volumn 48, Issue 1-3, 2002, Pages 253-285

  Collins, Michael ^a   Schapire, Robert E ^a   Singer, Yoram ^b  

^a AT AND T LABS RESEARCH   (United States)

^b HEBREW UNIVERSITY OF JERUSALEM   (Israel)

Author keywords

AdaBoost; Boosting; Bregman distances; Convex optimization; Information geometry; Iterative scaling; Logistic regression; Maximum entropy methods

Indexed keywords

ADA (PROGRAMMING LANGUAGE); ALGORITHMS; COMPUTATIONAL GEOMETRY; ITERATIVE METHODS; OPTIMIZATION; PROBLEM SOLVING; THEOREM PROVING;

BREGMAN DISTANCES; INFORMATION GEOMETRY; ITERATIVE SCALING; LOGISTIC REGRESSION;

REGRESSION ANALYSIS;

EID: 0036643072     PISSN: 08856125     EISSN: None     Source Type: Journal    
DOI: 10.1023/A:1013912006537     Document Type: Article

References (30)

1. A maximum entropy approach to natural language processing
2. The relaxation method of finding the common point of convex sets and its application to the solution of problems in convex programming
3. Arcing the edge
4. Prediction games and arcing classifiers
5. An iterative row-action method for interval convex programming
6. Bounds on approximate steepest descent for likelihood maximization in exponential families
7. A simple, fast, and effective rule learner
8. Why least squares and maximum entropy? An axiomatic approach to inference for linear inverse problems
9. Generalized projections for non-negative functions
10. Generalized iterative scaling for log-linear models
11. Inducing features of random fields
12. Duality and auxiliary functions for Bregman distances
13. Scaling up a boosting-based learner via adaptive sampling
14. Potential boosters?
15. The alternating decision tree learning algorithm
16. A decision-theoretic generalization of on-line learning and an application to boosting
17. Additive logistic regression: A statistical view of boosting
18. Robust trainability of single neurons
19. Additive versus exponentiated gradient updates for linear prediction
20. Boosting as entropy projection
21. Relative loss bounds for multidimensional regression problems
22. Additive models, boosting and inference for generalized divergences
23. Statistical learning algorithms based on Bregman distances
24. On-line learning of linear functions
25. Functional gradient techniques for combining hypotheses
26. Soft margins for AdaBoost
27. Improved boosting algorithms using confidence-rated predictions
28. BoosTexter: A boosting-based system for text categorization
29. From computational learning theory to discovery science