Homework answers / question archive / FIN650/Kim Project 3 Assignment Due Your goal is to develop a deep neural network algorithm with the smallest prediction error, i

FIN650/Kim Project 3 Assignment Due Your goal is to develop a deep neural network algorithm with the smallest prediction error, i

Statistics

Share With

FIN650/Kim

Project 3 Assignment Due

Your goal is to develop a deep neural network algorithm with the smallest prediction error, i.e., the model with the smallest root mean square error (RMSE) value. 2. Pick a stock of your team’s interest. Try to pick a stock that has been around for a long time – in general, the more data you feed into your model, the sharper its prediction becomes. 3. Modify the code to improve the model's performance. For example: a. Try to increase the sample period of your dataset. b. Try to change the percentage of data used for training. c. Modify the look-back period. d. Try using the GRU (Gated Recurrent Unit) model in lieu of the LSTM model. e. Try to add or drop the number of layers f. Try a different number of units in the layer. g. Try different batch sizes or epochs. i. Note that if you increase the number of epochs, it will take longer to train your model. h. Try to explore any other creative ways to improve your model. i. However, do not modify the dropout rate: