Time: 8 weeks

Team: 2

Language: Python

Since 2002, trading robots, AKA algorithmic trading, are considered to be one of the major breakthroughs on financial markets over the last decade. In concrete terms, these bots’ mission is to evaluate a share’s quotation, and to react by deciding to either buy or sell, through an automated process.These trading bots are able to compute thousands of orders per second (this is called High Frequency Trading) which in return, of course, modifies the very nature of markets. They are not merely trading spots anymore, but have become fighting arenas for various devices of artificial intelligence. Besides, this very promising field is still widely open; which is why the best scientists in the world are working on it. You are one of them.

You have to create a trading bot that will bring about a revolution of the trade markets.

The program ai-bot-workspace is the tool you’ll use to test your Trade project. Here’s a short guide to get you started.

• Download ai-bot-workspace v2.4.3 from GitHub

On Linux: ai-bot-workspace-2.4.3-x86_64.AppImage

There are two ways to install the application:

1. Go to the Releases page and download the latest version for your operating system.
2. Install Node.js and Yarn, clone this repository, cd into it and install the dependencies by running yarn. Start the application using yarn start.

• Create the binary

>> make

• Set the bot

Go to File->Settings In "Command[...]" field, provide the path to the previously generated binary called trade.

• Set datasets

A default datasets is provided. You can otherwise give a path to your own dataset in the "Location of the .csv file with candle data. [...]" field.

>> make tests_run

Author Corentin COUTRET-ROZET and PATRICIA MONFA-MATAS