Negative Log Likelihood Ratio Loss for Deep Neural Network Classification
Machine Learning
2018-05-01 v1 Machine Learning
Abstract
In deep neural network, the cross-entropy loss function is commonly used for classification. Minimizing cross-entropy is equivalent to maximizing likelihood under assumptions of uniform feature and class distributions. It belongs to generative training criteria which does not directly discriminate correct class from competing classes. We propose a discriminative loss function with negative log likelihood ratio between correct and competing classes. It significantly outperforms the cross-entropy loss on the CIFAR-10 image classification task.
Keywords
Cite
@article{arxiv.1804.10690,
title = {Negative Log Likelihood Ratio Loss for Deep Neural Network Classification},
author = {Donglai Zhu and Hengshuai Yao and Bei Jiang and Peng Yu},
journal= {arXiv preprint arXiv:1804.10690},
year = {2018}
}