Stock Trend and Price Prediction using Deep Learning Model (Using a sequence Model "LSTM: Long Short Term Memory Network")

Prediction of the stocks trend and closing price

Introduction About the Data :

The dataset We are directly fetching from the yfinance API. We can fetch any stocks data using their specific STOCK TICKER NAME (eg Apple:AAPL', Netflix: NFLX)

This data set from the yfinance consists of the following columns.

• Date :
• Open :
• High :
• Low :
• Close :
• Adj Close :
• Volume :

Target variable: Our target is to predict the Closing Price of the stock and also to predict the trend of the stock.

Approach for the project

1. Data Ingestion :

   • In Data Ingestion phase the data is first fetch from yfinance and coverted into the pandas dataframe
   • Plotted Some basic plotly and matplotlib charts for the closing price vs date along with some different moving averages 100 ma & 200 ma
   • Then the closing price data is split into training and testing and by following the time series principles.

2. Data Transformation :

   • In this phase a the data is transformed using MinMaxScaler.

3. Model Training :

   • LSTMS (A Sequential Model) is defined using keras sequential model for training.
   • For the regularization Dropout is defined
   • After finishing the training, trained model is saved in keras.h5 format

4. Prediction of the Time series :

   • Now model is utilized to predict the time series (sequential closing price of the stock)

5. Flask App creation :

   • Streamlit app is created with User Interface to predict stock trend and closing price inside a Web Application.

Model Training Approach Notebook

Link : Model Training LSTM Notebook

Streamlit WebApp Deployment:

Streamlit link : [https://stockpredlstm.streamlit.app/]

Screenshots of UI

Feeding the Stock Ticker Name and choosing the Time Frame:

Chart:

Predicted closing price vs Original closing price:

Evaluation Metric R2_Score: