🔮

Oracle™

An automated AI-backed stock prediction application.

Report Bug · Request Feature

Table of Contents

• About the Project
  • Built With
• Getting Started
  • Prerequisites
  • Installation
• Usage
• License
• Copyright

About The Project

Oracle™ is an application built to predict the movement of the stock market.

By leveraging state-of-the-art AI LSTM neural networks and vasts amount of data, Oracle™ learns which stock features signal a rise or fall in price.

Currently, it has about a 60% accuracy in predicting the daily movement of a stock.

Read the full academic research paper here.

Built With

• Python - The scripting language
• Anaconda - The scientific package management and deployment library
• Tensorflow - The machine learning library
• Keras - The high-level wrapper library over Tensorflow
• Numpy - The fast numeric library
• Pandas - The data frame library
• Scikit-learn - Another machine learning library

Getting Started

Follow these simple example steps to get a local copy up and running.

Prerequisites

• Anaconda

   https://www.anaconda.com/distribution/#download-section
  

Installation

On Windows, open Anaconda Prompt

On Mac/Linux, open terminal

1. Create the Oracle™ environment

   conda create -n oracle

2. Activate the Oracle™ environment

   conda activate oracle

3. Install Tensorflow (also installs Python and Numpy)

   • CPU (if no dedicated GPU)

      conda install tensorflow

   • GPU

      conda install tensorflow-gpu

4. Install Keras

   conda install keras

5. Install Pandas

   conda install pandas

6. Install Scikit-learn

   conda install scikit-learn

7. Clone the repo

   git clone https://github.com/ConnorLanglois/oracle.git

Usage

On Windows, open Anaconda Prompt

On Mac/Linux, open terminal

1. Activate the Oracle™ environment

   conda activate oracle

2. Navigate to src

   cd ~/.../oracle/src

3. Run the main file:

   python oracle.py

License

This project is intentionally not licensed.

Copyright

Copyright © 2019 Connor Langlois. All rights reserved.