Skip to content

Unsupervised machine learning models for the analysis of cryptocurrencies.

Notifications You must be signed in to change notification settings

abdulla971/Cryptocurrencies

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Cryptocurrencies

Overview

The main purpose of the challenge is to provide a report for an investment bank that is seeking to offer cryptocurrency investment portfolio. In the report the cryptocurrencies are shown on the trading market, and they are grouped to create a classification system this is done for the new investment. To perform the tasks for this project, I will perform unsupervised machine learning functions on data provided by CryptoCompare.

Results

I cleaned the data to keep actively traded cryptocurrencies, and that also a defined algorithm, also it must have a complete set of data. With all theses standards set only 532 different cryptocurrencies remained. After that I created a three dimensional graph that represents the groups of the different cryptocurrencies. For each point it has a name and the algorithm used to create the currency.

Then I added a two dimensional graph that represents the relationship between the total coin supply and the total coins mined this was done to compare each currency to the rest, also it is well noted that each point has its currency name.

Summary

Looking at the three dimensional graph, we observe that there is four different groups. There is two clear groups that are close to each other with almost all currencies falling within these two groups. One other group has a small number of currencies that act similarly, with few outliers. It is well noted that the outliers can be either over performers and under performers, however this needs further analysis to determine which is which. To do that I think of examining the total coin supply vs total coins mined graph. For the two main groups have the data mainly between 0% and 40% of the largest currency that is based on volume. It is noted that the group with the fewest currencies is placed close to 0% of the largest currency., on the other hand the group with one currency is at 100% since it is the largest,

Further analysis is required on the cryptocurrencies such as looking at their historical pricing to further understand how each currency performs. The would further help the investors with assessing the risks involved with the different cryptocurrencies.

About

Unsupervised machine learning models for the analysis of cryptocurrencies.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published