In this DaScient Capital proprietary program we apply Reinforcement Learning ensembles, the SuperTrend indicator, Kalman Filter Forecast Model/Parameterization, Simple & Exponential Moving Average(s), Relative Strength Index (RSI), as well as, other robust technical indicators that can be forked in template to calculate intelligent trade signals and suggests profitable buy & sell orders on both Robinhood, Binance.US, as well as TD Ameritrade asset/cryptocurrency trading platforms using Python.
You are welcome & encouraged to extrapolate, manipulate, and enhance your forked versions, as the elementary content of this project (i.e. decision science & forecast modeling) is most certainly applicable to any predictive/time-series/ml programs & efforts external to #trading & #assetmanagement. All that we ask is to cite our presence and respect the dignity of our paramount open-source community guidelines.
More details about initiation & project delevopment can be found in this Medium publication. This Kaggle Notebook showcases an automated solution with intricately calculated recommendations.
Current efforts are focused to include the incorporation of social sentiment analyses, more sophisticated technical indicators, and pre-built strategies to autonomously trade assets for users. Our platform aims to enable users to mirror peer portfolios. Users will be able to create their own or join publicly reviewed group-asset allocation pools rated by the vox populi. Additionally, we intend to assess & evaluate the bot's modeling parameters as they are autonomously hyper-tuned recurrently over time in such that decision signals are delivered with greater confidence the longer it runs.
Our ultimate goal is to optimize returns and minimize losses - for everyone. We want to make trading as easy and less intimidating as robotically possible.
Our mission is to leverage statistical & mathematical modeling, deep financial analytics, data science, strategic intelligence, and ml to provide reliable, intellectually resourceful, and totally open source user-friendly products.
Find value in our project? Want to help us expand or contribute to code? Even if it's just to say thanks - please feel free to contact us. We gladly only accept work contributions as donations, as this is clearly a work in progress, but do send a "hello, world!"
This special feature allows users to fully, or partially, imitate other user's portfolio allocations. The Trends enhancement allows users to become more exposed and informed of the SuperTrends trading community. Users can search openly for groups and strategies that they feel are more suitable for their level of risk. Users can also create, share, and update their own portfolios to the community as well, meaning whatever new allocations the group leader decides will change everyone's portfolios to reflect it! You can let your imagination wander.
The purpose of this repository is to document the development of our very first trading bot. The bot will be broken out into 3 broad classes:
- Data Class
- Trade Strategy Class
- Execution Class
Although our team is working very hard to get the fully user-friendly app developed & deployed, there will still be plenty of time to test the bot on your own as we work through our goals and plans for the future. Feel free to fork, star, and/or watch for any of our updates here on GitHub.
Here's what you'll currently need in order to execute the bot locally on your machine. (An introductory Python crash course probably wouldn't hurt.) If you do run with it we seek your inputs, suggestions, and ideas that you can prove have a place in our code. Whether it's to help the bot run more efficiently or how we can better scale our project. Your thoughts are welcome! This program is far from perfect, but your support and growing interest give us all the encouragement we need to get this released and trading as smoothly as we possibly can. Don't hesitate to send us an email if there's anything we can do to help: contact@dascient.com
- Python (Latest +3.9.7)
- Jupyter Notebook - Anaconda (mini-conda will certainly suffice)
- Binance.US crypto brokerage account. (API_KEY, API_SECRET)
- Lastly, you'll need this repository cloned somewhere easy to find. (i.e ./Desktop/GitHub will do.)
Don't have Binance.US? Sign up here!
- After login, go to the menu settings and find API MANAGEMENT.
- Create an API and follow approval directives.
- Save and KEEP ULTRA SAFE your api.key & api.secret (in a config.py, follow format).
Open terminal, locate binance_bot.py, then execute by typing:
$ python binance_bot.py
or whichever bot.py from folder of choice.
For Reinforcement Learning Mod - Binance Bot, follow these instructions.
This class will consist of a CCXT connection into Binance.US WebSocket interface that will feed live cryptocurrency data in the form of candle sticks; Open, High, Low, Close (OHLC).
We also apply rolling averages, upper/lower Bollinger bands, and binary variables that evaluates uptrend/downtrend intervals.
Like many things in life, sometimes one needs a little variety. There is no shortage of trade strategies to apply to our bot. With this in mind, the strategy class will be designed to be modular. That is, it is to be developed with "plug-and-play" design in order to develop different trading strategies over time. As long as the strategy sends a buy/sell signal for the execution, it will function properly.
Once the trade signal is sent, the execution class will send the order to Binance.US via ccxt.exchange.
Relax, have fun, and don't forget to drink plenty of water! 🎉🚀🌕
Part Time Larrys (hackingthemarkets)'s supertrend-crypto-bot