Readings

Loss Functions in Classification:

Group: Nan Ding, Youhan Fang

Statistical behavior and consistency of classification methods based on convex risk minimization. Tong Zhang.
Convexity, classification, and risk bounds. Peter Bartlett, Mike Jordan and Jon McAuliffe.
How to compare different loss functions and their risks. Ingo Steinwart.
On surrogate loss functions and f-divergences. X. Nguyen, M.J. Wainwright and M.I. Jordan. Annals of Statistics, (37)2, 876--904, 2009.

Rademacher Complexities

Group: Duncan Leaf, Sanvesh Srivastava, Yen-Ning Huang

Rademacher penalties and structural risk minimization. IEEE Transactions on Information Theory 47(5): 1902-1914 (2001). Vladimir Koltchinskii
Rademacher and Gaussian Complexities: Risk Bounds and Structural Results. Peter Bartlett, Shahar Mendelson. JMLR.
Local Rademacher Complexities Peter Bartlett, Olivier Bousquet, Shahar Mendelson.
Local Rademacher complexities and oracle inequalities in risk minimization. Vladimir Koltchinskii.

Boosting

Group: Cheng Liu, Yuan Lin

Process consistency for AdaBoost. Wenxin Jiang. Annal of Stat.
AdaBoost is Consistent. P. L. Bartlett and M. Traskin.
Some Theory for Generalized Boosting Algorithms. P. J. Bickel, Y. Ritov and A. Zakai.
Boosting with early stopping: Convergence and Consistency. T. Zhang and B. Yu.

Multiclass Classification

Group: Suleyman Certintas and Timothy La Fond

On the consistency of multiclass classification Ambuj Tewari and Peter Bartlett.
Multicategory SVM: Theory and Application to the Classification of Microarray Data and Satellite Radiance Data. Y. Lee, Y. Lin and G. Wahba.
An infinity sample theory for multi-category large margin classification. Tong Zhang.
Fisher Consistency of Multicategory SVM. Yufeng Liu.

Aggregation Learning

Group: Zhaonan Sun and Nawanol Theera-Ampornpunt

Aggregation for Gaussian regression. Florentina Bunea, Alexandre Tsybakov and Marten Wegkamp Annals of Statistics, Vol 35, 1674 - 1697 (2007).
Optimal Aggregation of Classifiers in Statistical Learning. A. Tsybakov. Annals of Statistics.
Aggregation and Sparsity via l1 penalized least squares. F. Bunea, A. Tsybakov and M. Wegkamp.

Penalty-based Model Selection

Risks bounds for model selection via penalization. Andrew Barron, Lucien Birge, Pascal Massart.
From model selection to adaptive estimation. L. Birge and P. Massart.
Minimum contrast estimators on sieves: exponential bounds and rates of convergenceL. Birge and P. Massart.