Stock Market Prediction using Deep Learning

The stock market prediction has been a focus for years since it can yield significant profit and can be one of the most rewarding experiences. But at the same time, predicting the stock market is one of the difficult things to do. There are so many factors involved which may affect stock prices. Also, another challenge in stock market prediction is volatility, stocks are one of the most volatile instrument.

There are different methods to predict future prices are:

  • Fundamental analysis
  • Technical analysis
  • Techno-Fundamental approach
  • Time series models – ARIMA
  • Deep learning models – LSTM (Long short term memory neural networks) – will be using LSTM for this web app.
What is LSTM?

LSTMs are used for sequence prediction problems. For example, LSTM is applicable to tasks such as unsegmented, connected handwriting recognition, speech recognition, Natural language problems, etc.

A common LSTM unit is composed of a cell, an input gate, an output gate and a forget gate. The cell remembers values over arbitrary time intervals and the three gates regulate the flow of information into and out of the cell. Due to these capabilities, these networks are well suited to make predictions on time-series data.

LSTM Cell
Webapp Capabilities
  • Create stock watchlist, add as many stocks to your watchlist
  • Quick snapshot to provide an easy view of basic fundamental information for all the stocks in the watchlist.
  • Download utility to load the EOD data from the yahoo API to the SQLLite database.
  • Train a deep learning model on the stocks.
  • Generate predictions and forecast prices for the next 4 days.
Technology Stacks
  • Python
  • Important packages – Tensorflow, Keras, NumPy, pandas, and Sklearn
  • Django
  • Chart.js/Highchart.js
  • Third-party APIs to fetch stock data
  • Database -SQLLite
Screenshots
Main Page

Train Deep Learning Model

Show Predictions for next 20 days

Leave a Comment

Your email address will not be published. Required fields are marked *