The purpose of this project is to create an analysis of cryptocurrency data for a client who is interested in investing in the cryptocurrency market. This analysis uses unsupervised machine learning to find patterns in the cryptocurrency market dataset in order for the client to select the best cryptocurrency in which to invest. In order to accomplish the analysis techniques in preprocessing data, clustering data and reducing dimensions within the dataset, and reducing principal components using PCA are demonstrated.
The purpose of unsupervised learning is to identify patterns and trends within the data with which to make decisions. The charts, table, and dataframe below are the products and visualizations this unsupervised machine learning analysis yielded.
This 3D plot shows the data after the number of dimensions were reduced through PCA (principal component analysis) and with the clusters applied by the unsupervised ML model
The analysis began with a set of cryptocurrencies which were all candidates for investment. This table shows the listing of tradable cryptocurrencies after the analysis is complete and is interactive as the individual columns are selectable.
This dataframe screenshot displays the first 10 cryptocurrencies after the data has been scaled and clustered by the ML model.
The scatter plot below displays the tradable cryptocurrencies graphically and they are color-coded based on the class the ML model assigned to them.