Einstock Algo Trader

A stock trading simulation app that uses machine learning algorithms to predict equity price trends.

Team

• Product Owner: Kevin Kim
• Scrum Master: Lucas Hawes
• Lead Frontend Engineer: Aaron Stevens
• Lead Backend Engineer: Natasha Che

Table of Contents

1. Features
2. Requirements
3. Development
   1. Installing Dependencies
   2. [Tasks] (#tasks)
4. Contributing

Features

Multiple algorithms

Popular machine learning algorithms, including K Nearest Neighbors, Logistic regressions, Naive Bayes, Neural Networks, and Random Forests, are available for testing

Wide range of equity selections

Users can test algorithm performance with historical data for over 7000 stocks traded on NYSE and Nasdaq

Hassle-free trade simulation

Buying and selling decisions are automatically generated according to the price predictions from the selected algorithm

Comprehensive evaluation

Einstock reports various assessment metrics after a simulation run to evaluate an algorithm's performance against market benchmarks

Requirements

• Node 6.1.x
• Postgresql 9.1.x

Development

Installing Dependencies

From within the root directory:

npm install
bower install

Tasks

Start the local database:

postgres -D /usr/local/var/postgres

Start the server:

npm start

Contributing

See CONTRIBUTING.md for contribution guidelines.