https://github.com/bishopce16/cryptocurrencies

An analysis on cryptocurrencies dataset using unsupervised machine learning, PCA algorithm, and K-means clustering.
https://github.com/bishopce16/cryptocurrencies

hvplot jupyter-notebook pandas plotly python scikit-learn unsupervised-machine-learning visual-studio-code

Last synced: 21 days ago
JSON representation

An analysis on cryptocurrencies dataset using unsupervised machine learning, PCA algorithm, and K-means clustering.

• Host: GitHub
• URL: https://github.com/bishopce16/cryptocurrencies
• Owner: bishopce16
• Created: 2022-10-08T09:11:47.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2022-10-10T21:07:39.000Z (over 2 years ago)
• Last Synced: 2024-11-10T02:25:53.636Z (3 months ago)
• Topics: hvplot, jupyter-notebook, pandas, plotly, python, scikit-learn, unsupervised-machine-learning, visual-studio-code
• Language: Jupyter Notebook
• Homepage:
• Size: 4.63 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# cryptocurrencies

An analysis on cryptocurrencies dataset using unsupervised machine learning, PCA algorithm, and K-means clustering.

## Overview of Project

The purpose of this project was to analyze a cryptocurrencies dataset using unsupervised machine learning, PCA algorithm, and K-means clustering. To create a report and visualization of currently traded cryptocurrencies categorized by grouping them by their features, for a new investment portfolio on cryptocurrencies.

---

## Resource:

Data Sources: crypto_data.csv

Tools: Visual Studio Code, Jupyter Notebook, Python, Unsupervised Machine Learning, pandas, Scikit-learn, Ploty, hvPlot

 

---

## Results and Summary:

The dataset crypto_data.csv was retrieved from [CryptoCompare ](https://min-api.cryptocompare.com/data/all/coinlist), containing 1,252 entries. Only 1,144 of the cryptocurrencies were currently trading, once the null values were removed. Just cryptocurrencies that had a total number of mined coins greater than zero remained, leaving 532 tradable cryptocurrencies.

![ crypto_df_drop_coinname.png](images/crypto_df_drop_coinname.png)

 Tradable cryptocurrencies table

![ clustered_df_hvplot_table.png](images/clustered_df_hvplot_table.png)

Using the K-means algorithm, an elbow curve found the best k-value seems to be k=4. Settling the cryptocurrencies would have an output of 4 clusters to be categorized.

![ elbow_curve_hvplot.png](images/elbow_curve_hvplot.png)

The 3-D scatter plot below made by reducing the cryptocurrencies to three principal components using the PCA algorithm.

![3d_scatter_wclusters.png](images/3d_scatter_wclusters.png)

The 2-D scatter plot shows (TotalCoinsMined on the x-axis and TotalCoinSupply on the y-axis)  the cryptocurrencies distribution of the 4 clusters. 

![ plot_df_hvplot_scatter.png](images/plot_df_hvplot_scatter.png)

---