Hinge Loss

Hinge loss is a crucial component of Support Vector Machines (SVMs) and other machine learning models, primarily used in binary classification tasks to maximize the margin between classes while minimizing the impact of outliers. Current research focuses on improving the efficiency and robustness of hinge loss-based methods, exploring variations like p-norm hinge loss and smoothed versions to address challenges such as computational cost and sensitivity to noise in large datasets. These advancements, along with investigations into the theoretical limits of hinge loss in deep learning architectures and its application in multi-task learning and hierarchical classification, aim to enhance the accuracy and scalability of classification models across diverse applications.

Papers

August 19, 2024

$p$SVM: Soft-margin SVMs with $p$-norm Hinge Loss
Haoxiang Sun
Support Vector Machine Soft Margin Hinge Loss

December 17, 2023

A Smoothing Algorithm for l1 Support Vector Machines
Ibrahim Emirahmetoglu, Jeffrey Hajewski, Suely Oliveira, David E. Stewart
Support Vector Machine Soft Margin Hinge Loss Smoothing Algorithm Norm Penalty

September 13, 2023

Fundamental Limits of Deep Learning-Based Binary Classifiers Trained with Hinge Loss
Tilahun M. Getu, Georges Kaddoum
Neural Network Deep Learning Binary Classifier Fundamental Limit Hinge Loss Performance Limit

August 31, 2023

Majorization-Minimization for sparse SVMs
Alessandro Benfenati, Emilie Chouzenoux, Giorgia Franchini, Salla Latva-Aijo, Dominik Narnhofer, Jean-Christophe Pesquet, Sebastian J. Scott, Mahsa Yousefi
Support Vector Machine Majorization Minimization Hinge Loss Sparse Classification

June 24, 2023

Kernel Support Vector Machine Classifiers with the $\ell_0$-Norm Hinge Loss
Rongrong Lin, Yingjia Yao, Yulan Liu
Support Vector Machine Hinge Loss Kernel Support Vector Machine

June 16, 2023

Training shallow ReLU networks on noisy data using hinge loss: when do we overfit and is it benign?
Erin George, Michael Murray, William Swartworth, Deanna Needell
Training Data Gradient Descent Binary Classification Noisy Data Model Overfitting Two Layer ReLU Benign Overfitting Shallow ReLU Hinge Loss

December 26, 2022

Improved Kernel Alignment Regret Bound for Online Kernel Learning
Junfan Li, Shizhong Liao
Kernel Matrix Kernel Alignment Hinge Loss Online Kernel

December 24, 2022

Iterative regularization in classification via hinge loss diagonal descent
Vassilis Apidopoulos, Tomaso Poggio, Lorenzo Rosasco, Silvia Villa
Machine Learning Classification Code Hinge Loss Iterative Regularization

October 31, 2022

Xtreme Margin: A Tunable Loss Function for Binary Classification Problems
Rayan Wali
Loss Function Binary Classification Margin Maximization Surrogate Loss Hinge Loss

October 19, 2022

Hierarchical classification at multiple operating points
Jack Valmadre
Simple Classifier Hierarchical Classification Softmax Classifier Node Prediction Hinge Loss Top Down Supervised Learning Approach

October 12, 2022

Optimizing Evaluation Metrics for Multi-Task Learning via the Alternating Direction Method of Multipliers
Ge-Yang Ke, Yan Pan, Jian Yin, Chang-Qin Huang
Multi Task Learning Generalization Performance Evaluation Metric Class Wise Multiplier Hinge Loss

May 24, 2022

Soft-SVM Regression For Binary Classification
Man Huang, Luis Carvalho
Loss Function Novel Regression Binary Classification Logistic Regression Soft Margin Hinge Loss

February 4, 2022

Stochastic smoothing of the top-K calibrated hinge loss for deep imbalanced classification
Camille Garcin, Maximilien Servajean, Alexis Joly, Joseph Salmon
Dynamic Loss Hinge Loss Node Aware Bi Smoothing Deep Imbalanced