this project is aimed for training a deep neural network model to predict exchange rate.
As of now I have design a simple DNN model to generate predictions. The input of my model is the exchange rate result from the last 30 days. The output of my model would be the prediction. There are 2 hyperparameter(epoch,learning rate)for this model.
The next modification for this project would be upload soon. It will include more features:
user can decide which kind of exchange rate to predict on a user-friedly interface, and how many days' exchange rate data as input of the prediction model(last 30 days by default).
the loss of the training process will be plotted, and the relation between each input day with the final prediction will be displayed.
user can modify the structure of neural network if they want.