/Stock-Prediction

Primary LanguageJupyter Notebook

A Beginner Approach to Predicting Stock Closing Price using Long Short-Term Memory Models

For our final project in STAT 430: FDL, we will construct an Artificial Neural Network that can accurately recognize an increasing or decreasing trend in market data, and predict the closing price for the following day. This will be achieved through the use of the plethora of financial data available for free to the general public. Many websites provide daily quote data for individual stocks that date back a varying number of years. We have narrowed down the stocks we will be using to Apple Inc. (AAPL), Google LLC (GOOGL),Amazon Inc. (AMZN), and Microsoft Corporation (MSFT). The majority of this data will be used to train our model and the remainder will be used to test it. Each observation used to train and test this model will be the daily closing price of the respective stock. The closing price is the last price the stock was at before the market closed for the day. The input neurons in this model will contain the aforementioned closing prices,and will be the only data used to train this network. Our expected output is the predicted closing price for the next day.