SOme USeful NOtes for MYself.
SOme USeful NOtes for MYself.
B站神奇的频道(YouTube里同名):关于微积分/线代/梯度下降/DL等数学知识的理解,对理解DL很有帮助
https://space.bilibili.com/88461692/#/channel/detail?cid=9450
知乎有用的总结:
https://zhuanlan.zhihu.com/p/55519131?utm_source=qq&utm_medium=social&utm_oi=1101546246992474112
100天ML-Code教程:旨在提高Codding能力
https://github.com/MLEveryday/100-Days-Of-ML-Code
英文原版:https://github.com/Avik-Jain/100-Days-Of-ML-Code
目标检测算法的学习路线(全),涵盖了几乎所有目标检测算法
https://github.com/hoya012/deep_learning_object_detection
凸优化主流算法归纳
https://zhuanlan.zhihu.com/p/47453144?utm_source=qq&utm_medium=social&utm_oi=1101546246992474112
一个神奇的wechat公众号:超智能体。同作者B站:YJango。讲述了很多有用的数学/思考知识,专题“学习观”。熵的概念几乎是全网最清晰的讲解。讲述了深度学习发展以及各个方向。
B站地址:https://space.bilibili.com/344849038?from=search&seid=9485701162385222147
从泰勒展开来看梯度下降:
star数很高的印度小哥创的仓库,python练习,算法练习等,推荐
https://github.com/TheAlgorithms/Python
复旦大学邱锡鹏老师的书
https://zhuanlan.zhihu.com/p/61618061?utm_source=qq&utm_medium=social&utm_oi=1101546246992474112
github:https://github.com/nndl/nndl.github.io
配套邱老师的练习
https://github.com/nndl/exercise/tree/master/warmup
示例代码:https://link.zhihu.com/?target=https%3A//github.com/nndl/nndl-codes
免费GPU(虽然kaggle-kernal不好用)
https://zhuanlan.zhihu.com/p/59305459?utm_source=qq&utm_medium=social&utm_oi=1101546246992474112
如何画神经网络模型
Kaggle入门
https://zhuanlan.zhihu.com/p/25686876
https://zhuanlan.zhihu.com/p/27424282
https://www.cnblogs.com/limitlessun/p/8489749.html
特征工程详解
https://www.zhihu.com/question/29316149
适合入门的比赛:
Predict future sales:https://www.kaggle.com/c/competitive-data-science-predict-future-sales/data