Welcome to the fourth programming assignment of the Deep Learning course.
The goal is to predict stock price. Prediction stock price is a time-series problem. Time series prediction problems are a difficult type of predictive modeling problem.
Unlike regression predictive modeling, time series also adds the complexity of a sequence dependence among the input variables.
A powerful type of neural network designed to handle sequence dependence is called recurrent neural network. The LSTM and GRU networks are a type of recurrent neural network used in deep learning because they can successfully train very large architectures.
After this assignment, you will:
- Discover how to develop LSTM/GRU networks to address a demonstration time-series prediction problem.
- Have a comparison between LSTM and GRU.
Google Stock Dataset This dataset consists of the closing stock price of a share of Google stock during the trading days between December 20, 2013, and December 17, 2021.link
GOOG.close The closing stock price of a share of Google stock.
GOOG.close is used for this problem.
-- Create the Training Data in the Many-To-One relationship for LSTM/GRU (creating inputs and output for model training)
-- Train different LSTM/GRU models to predict stock price. (predicting `close` feature)
-- Compare the resluts for LSTM and GRU. (plotting the results in a graph besides comparing the metrics)
> Consider future predictions for the next 30 days. (30 time steps)