In this project, an analysis has been done to find out what cryptocurrencies are on the trading market and how cryptocurrencies could be grouped toward creating a classification for developing a new investment product.
Python, sklearn, K-means algorithm, PCA algorithm, Pandas, hvplot, plotly
-
Since there was no known output for what we are looking for, unsupervised learning has been used.
-
The Cryptocurrencies data given was not ideal, so it was preprocessed and scaled to fit the machine learning model.
-
The data dimensions were reduced to three principal components using PCA algorithm from sklearn.
- Clusters of Cryptocurrencies data were predicted by plotting the elbow curve to find the best value for K using KMeans algorithm. By looking at the elbow curve, the best K value was found 4. The model was initialized with 4 clusters using KMeans algorithm. The fitting and prediction of data were made based on this model.
- Data-Table, 3D-scatter and 2D-scatter plots have been used to show the status of the current tradable cryptocurrencies with all the relevant informations.
- From 2D-scatter plot, it can be seen that there are some outliers like "BitTorrent" and "TurtleCoin"