Udemy: deep learning A-Z
This RNN model is trained on 5 years of google stock prices. The goal is to predict the January 2017 google stock price trends.
The model is trained on only the past 5 years, not on January prices. The actual January google stock prices are used for comparison purposes.
At the end, the predicted results and the actual results for January 2017 is compared. The goal of this model is to predict the trends (upward or downward) of the google stock price, instead of predicting the price itself.
The plot at the end of the notebook shows that this model was succesful to capture the trends