Alex Krizhevsky
通过这篇论文，一定程度上了解了deep learning 和image recognition 的结合
亮点：
1，神经元输出的调节函数，弃用sigmoid和tanh，转用nonlinearity方法： Rectified Linear Units（ReLUs）
2，多个GPU 共建一个神经网络时候的小trick
3，local normalization scheme 来帮助泛化
4，避免overfitting的两个方法
（1）data augmentation
（2）dropout
一．论文概述 Alex Krizhevsky、Ilya Sutskever和Geoffrey Hinton训练了一...
Alex Krizhevsky 通过这篇论文，一定程度上了解了deep learning 和image recog...
AlexNet (ImageNet Classification with Deep Convolutional ...
原文：ImageNet Classification with Deep ConvolutionalNeural ...
ImageNet Classification with Deep Convolutional NeuralNet...
原文：《ImageNet Classification with Deep Convolutional Neura...
ImageNet Classification with Deep Convolutional Neural Ne...
ImageNet Classification with Deep Convolutional Neural Ne...
ImageNet Classification with Deep Convolutional Neural Ne...
ImageNet Classification with Deep Convolutional Neural Ne...
本文标题：《ImageNet Classification with De
本文链接：https://www.haomeiwen.com/subject/ohilattx.html
网友评论