Skip to content

cuixuage/2018SpringCS231n

Repository files navigation

课程总结
通过课程的notes + slides学习的个人总结;查看0_CS231n_Course.md
参考链接:
课程主页:http://cs231n.stanford.edu/index.html
中文版notes: https://zhuanlan.zhihu.com/p/21930884

作业
assignment1
几种常见分类器:
1.KNN
2.linear-SVM
3.softMAx
4.TwolayerNet fully-connected

assignment2
1.Fully-connected任意深度NN
2.batch-normalize 这部分我没有做
3.dropout 概率P
4.CNN实现 这部分我没做
5.Pytorch nn.moudle optimizer 等等

assignment3
1.RNN实现 这部分我没做 demo:image-capation
2.LSTM长短期记忆 这部分我没做
3.style-transform 用的squeezenet 但是我不是很理解 跳过
4.GAN 关键是Loss的计算 deep-convolutional-GAN效果不错 demo:手写数字image生成
参考链接:
https://github.com/haofeixu/standford-cs231n-2018
https://github.com/lightaime/cs231n

Releases

No releases published

Packages

No packages published