Abstract:This paper examines the role and efficiency of the non-convex loss functions for binary classification problems. In particular, we investigate how to design a simple and effective boosting algorithm that is robust to the outliers in the data. The analysis of the role of a particular non-convex loss for prediction accuracy varies depending on the diminishing tail properties of the gradient of the loss -- the ability of the loss to efficiently adapt to the outlying data, the local convex properties of the loss and the proportion of the contaminated data. In order to use these properties efficiently, we propose a new family of non-convex losses named $\gamma$-robust losses. Moreover, we present a new boosting framework, {\it Arch Boost}, designed for augmenting the existing work such that its corresponding classification algorithm is significantly more adaptable to the unknown data contamination. Along with the Arch Boosting framework, the non-convex losses lead to the new class of boosting algorithms, named adaptive, robust, boosting (ARB). Furthermore, we present theoretical examples that demonstrate the robustness properties of the proposed algorithms. In particular, we develop a new breakdown point analysis and a new influence function analysis that demonstrate gains in robustness. Moreover, we present new theoretical results, based only on local curvatures, which may be used to establish statistical and optimization properties of the proposed Arch boosting algorithms with highly non-convex loss functions. Extensive numerical calculations are used to illustrate these theoretical properties and reveal advantages over the existing boosting methods when data exhibits a number of outliers.

An Efficient Version of Inverse Boosting for Classification

Imbalanced Data Sets Classification Method Based on Over-Sampling Technique

A Technique For Improving The Performance Of Naive Bayes Text Classification

Partially Corrective AdaBoost

Hybrid SVM algorithm oriented to classifying imbalanced datasets

RotBoost: A technique for combining Rotation Forest and AdaBoost

Improved Boosting Algorithm With Adaptive Filtration

A Direct Approach to Multi-class Boosting and Extensions

{\Cal U}Boost: Boosting with the Universum

Boosting in the presence of outliers: adaptive classification with non-convex loss functions

The success of AdaBoost and its application in portfolio management

An improved floatboost algorithm for Naïve bayes text classification

Bhattacharyya boosting

W-Boost and Its Application to Web Image Classification

A Method to Boost Naïve Bayesian Classifiers

Boosting Through Optimization of Margin Distributions

BalancedBoost: A hybrid approach for real-time network traffic classification

An Improved FloatBoost Algorithm for Naïve Bayes Text Classification.

Experimental Study and Comparison of Imbalance Ensemble Classifiers with Dynamic Selection Strategy

A hybrid ensemble and evolutionary algorithm for imbalanced classification and its application on bioinformatics

A New Diverse AdaBoost Classifier