BriefGPT.xyz
Jul, 2021
正则化分类感知量化
Regularized Classification-Aware Quantization
HTML
PDF
Daniel Severo, Elad Domanovitz, Ashish Khisti
TL;DR
本文提出了一种正则化分类感知量化的算法，该算法通过在重构误差上进行正则化0-1损失，用于二元分类任务，性能比文献中的基准量化方案更优，并能在未见数据上快速处理。
Abstract
Traditionally,
quantization
is designed to minimize the reconstruction error of a data source. When considering downstream classification tasks, other measures of distortion can be of interest; such as the 0-1
classific
→