Update homework9.ipynb

homeworks/homework9.ipynb

@@ -46,7 +46,7 @@
  "\n",
  "## 3. Time Series Implementation\n",
  "\n",
- "1. Implement a model similar to `RNNModel` of section [d2l.ai/chapter_recurrent-neural-networks/rnn-gluon.html](https://en.d2l.ai/chapter_recurrent-neural-networks/rnn-gluon.html) suitable for regression. It should take as input vector-valued data, such as the time series mentioned above and it should output vector-valued data (of some other dimensionality).\n",
+ "1. Implement a model similar to `RNNModel` of section [d2l.ai/chapter_recurrent-neural-networks/rnn-concise.html](https://en.d2l.ai/chapter_recurrent-neural-networks/rnn-concise.html) suitable for regression. It should take as input vector-valued data, such as the time series mentioned above and it should output vector-valued data (of some other dimensionality).\n",
  "1. Train the model on the first 4 years of data using plain RNN, GRU and LSTM cells (for a single layer). How well can the model \n",
  " * Predict the stock value the next day on the last 1 year of data (price at opening).\n",
  " * Plot how the quality of the model degrades as we apply it throughout the year (i.e. we ingest all the data up to day $t$ and predict forward at day $t+1$). \n",