[Reading] Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
1. 概述 Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift 一文提交于 2015/2，文章提出了一种通过对网络各层输入的每个小批量（mini-batch）进行标准化，来解决内部协变量偏移（internal covariate shift）的方法，称为批量标准…
Read more