💸 Machine Learning approach to profit on cryptocurrency markets 💸
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.ipynb_checkpoints
public
.gitignore
README.md
Untitled.ipynb
charts_info.py
coinmarket.py
giddy.py
gimmeDatDATA.py
main.py
plot.py
prices.py
requirements.txt
roundTwo.py
simplistic.py

README.md

#Coomb

###Concept

I tried to programmatically buy and sell different cryptocurrencies using machine learning to make money from the fluctuating currency values. This website does a good job of displaying the different cryptocurrencies and their prices over the last 7 days.

###How it works

Coomb uses the UCB1 Multi Arm Bandit Algorithm to process inputs of data and graph the potential of reward corresponding with each "arm" (option)!

Based on this work: https://www.quantopian.com/posts/multi-armed-bandit

Here were the results that I got from 9 hours of data aggregation: alt text