正文
1.【few-shot表示学习】《Learning to represent tasks for few-shot learning (Communication, Part 1) | ARAYA》by Nicholas Guttenberg
http://www.araya.org/archives/1971
GitHub：
https://github.com/arayabrain/FewshotClassifier
pdf：
https://pan.baidu.com/s/1qYI0p6c
@爱可可-爱生活
2.谷歌输入法背后的机器智能：思你所思，想你所想！
https://www.leiphone.com/news/201705/epzHTYkbjmABrv6H.html
By 雷锋网
3.【OpenAI Baselines：DQN】OpenAI基线：DQN。OpenAI开放OpenAI基线，其内部努力重现的强化学习算法，性能与已发表的结果相当。 OpenAI基线将在未来几个月发布算法; 今天的版本包括DQN及其三个变体。
https://blog.openai.com/openai-baselines-dqn/
@网路冷眼
4.【RNN详解】《Unfolding RNNs》by Suriyadeepan Ram
Part1:Concepts and Architectures
http://suriyadeepan.github.io/2017-01-07-unfolding-rnn/
Part2:Vanilla, GRU, LSTM RNNs from scratch in Tensorflow
http://suriyadeepan.github.io/2017-02-13-unfolding-rnn-2/
github：
https://github.com/suriyadeepan/rnn-from-scratch
@爱可可-爱生活
5.【贝叶斯：深度网络泛化的秘密】《Everything that Works Works Because it's Bayesian: Why Deep Nets Generalize?》
