Skip to content

Gather historical stock price data together with data from Google trends and train neural networks with fast.ai

Notifications You must be signed in to change notification settings

daMichaelB/stock-price-prediction

Repository files navigation

Stock Price Data - Downloader and Deep Learning

This repository delivers all required parts for Deep Learning on Stock Data

Stock data management

  • Stock data downloader
  • Stock data visualization with Streamlit
  • Stock data handling on local *.csv files

Prerequisites

Create a virtual environment and install required packages

virtualenv venv
pip install -r requirements.txt

Get your API-Key from Alphavantage

Go to Alpha Vantage and create your API-KEY.

Set the API-Key as environment variable:

export API_KEY="YourAPIKey"

Usage:

  • streamlit run download_stock_data.py
  • Select the stocks that should be downloaded and visualized
  • There already is a huge list of stock symbols that can be easily extended

You should see the following page in your browser:

stock analyzer

Google Trends data

  • Downloader for Google Trends
  • Interactive Google Trends visualization with Streamlit
  • Google Trends data handling on local *.csv files

Usage:

  • streamlit run download_google_trends.py
  • Select the stocks for which the trend should be downloaded and visualized
  • There already is a huge list of stock symbols that can be easily extended

Deep Learning with fast.ai

With stock data and google trend data downloaded in the previous steps, you can now train a deep neural network for a specific stock.

Usage:

  • streamlit run train_lstm_model.py

Based on Stock-Data and Google-Trend-Data a neural network (customized LSTM) is trained to predict future stock prices.

About

Gather historical stock price data together with data from Google trends and train neural networks with fast.ai

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published