This project examines the performance of the k-Means unsupervised learning model on a data set containing various metrics on several cryptocurrencies.
Prior to training and running the model, the data is cleaned, scaled, and then dimensionally reduced using PCA and t-SNE.
All code and output are contained in the Jupyter notebook: "crypto-cluster.ipynb".
Conclusions about the performance of this model and the possibility of this data being clusterable are also contained within the notebook.
To examine the data and code files, their locations are indicated in the tree below.
- "Code" (folder)
- "crypto-cluster.ipynb" (program for cleaning and modeling the cryptocurrency data and measuring it's performance)
- "crypto_data.csv" (cryptocurrency data file used by the program above)
(Please do not delete, move, rename, or alter!)