This is the code for a 4-layered recurrent neural network that predicts the trends in Google stock price based on previous 5 years of Stock data. I built the model using the Keras library, which is built on top of Tensorflow and Theano. The inputs are numeric values of opening and closing stock price for 5 years from 2012 to 2016. The prediction is done for the Opening stock value for the month of January 2017. I used adam for stochastic optimization, and mean_squared_error as the loss function.
● tensorflow ● keras ● numpy ● pandas ● scikit-learn
| Variable | Definition |
|---|---|
| Date | Date of stock price record |
| Open | Opening value of stock price |
| High | Highest value of stock price on that day |
| Low | Lowest value of stock price on that day |
| Close | Closing value of stock price |
| Volume | Total Volume of trade |
Run Google_Stock_Price.py in terminal to see the network in training. I have used Spyder from Anaconda to script and visualize the code.
Epoch 95/100
1198/1198 [==============================] - 7s - loss: 0.0015
Epoch 96/100
1198/1198 [==============================] - 7s - loss: 0.0014
Epoch 97/100
1198/1198 [==============================] - 7s - loss: 0.0015
Epoch 98/100
1198/1198 [==============================] - 7s - loss: 0.0016
Epoch 99/100
1198/1198 [==============================] - 7s - loss: 0.0013
Epoch 100/100
1198/1198 [==============================] - 7s - loss: 0.0014
Steps that can be taken to further improve the model:
- Getting more training data: We trained our model on the past 5 years of the Google Stock Price but it would be even better to train it on the past 10 years.
- Increasing the number of timesteps: the model remembered the stock prices from the 60 previous financial days to predict the stock price of the next day. That’s because we chose a number of 60 timesteps (3 months). You could try to increase the number of timesteps, by choosing for example 120 timesteps (6 months).
- Adding some other indicators: if you have the financial instinct that the stock price of some other companies might be correlated to the one of Google, you could add this other stock price as a new indicator in the training data.
- Adding more LSTM layers: we built a RNN with four LSTM layers but more layers can be added depending on buisness intuition.
- Adding more neurons in the LSTM layers: we highlighted the fact that
we needed a high number of neurones in the LSTM layers to respond better
to the complexity of the problem and we chose to include 50 neurones in
each of our 4 LSTM layers. You could try an architecture with even more neurones in each of the 4 (or more) LSTM layers.