# 201909. Week7
###### tags: `ML讀書會`

# 本週作業
1. 上周意識到直接講LSTM有點太難，感覺大家還並未熟悉這塊DL，所以線上課程主要以教大家如何應用LSTM為主
* 多層LSTM 模型預測
* Uber's 流量預測所使用的RNN架構
3. 若對LSTM理論有興趣的朋友，相關文件列在下面，有問題再來找我討論囉~
* [201909W7線上課程連結](https://www.youtube.com/watch?v=Lkcxtf2AZzQ&feature=youtu.be)

* [201909W7線上課程code連結](https://colab.research.google.com/drive/1yw_X3DU8KtW5-YKlpz_rrz1YlKhsY9qS)

5. [實機作業 & Q&A]
* 本週實機作業，meeting時間公布

6. Reference
* [lstm autoencoder ](https://machinelearningmastery.com/lstm-autoencoders/)
* [lstm autoencoder2 ](https://towardsdatascience.com/step-by-step-understanding-lstm-autoencoder-layers-ffab055b6352)
* [LSTM Model Architecture for Rare Event Time Series Forecasting paper](http://roseyu.com/time-series-workshop/submissions/TSW2017_paper_3.pdf)
* [LSTM Model Architecture for Rare Event Time Series Forecasting blog](https://eng.uber.com/neural-networks/)
* [lstm理論內容１](http://colah.github.io/posts/2015-08-Understanding-LSTMs/)
* [lstm理論內容２](https://www.zybuluo.com/hanbingtao/note/581764)