This is an analysis of various cryptocurrencies for an investment bank that wants to offer the cryptocurrencies as an additional form of investment. Machine learning was used for the analysis, and since the the bank didn't exactly know what outputs they were looking for, we have used unsupervised learning. The provided in a csv file data needed some preparations and cleaning in order for it to fir the machine learning models. To demonstrate that the cruptocurrncies can be grouped, we have used a clustering alghorythm and we have also used hvplot and plotly to visualize it.
- Google Colab (used because hvplot doesn't work on my computer)
- Skikit learn KMeans, StandardScaler, MinMaxScaler, PCA
- Hvplot, Plotly.
- Python (with Pandas, Numpy, Path, etc. dependecies
- Data from a csv file (available in this repository)
We have 532 various cryptocurrencies at hand. The four clusters can be seen in the 3D visual below. The 2D visual describes the relationship between every cryptocurrency from the csv file to overall coins mined and overall coin supply.
