Training data to make predictions for Crypto Investments
The popularity of bitcoin has caused a price jump that makes it unavailable for most investors. There are many other cryptocurrencies that are much more affordable. To sort through the mass of options out there I will use Machine Learning to make informed investment choices.
Before I begin, I consider the following questions:
- What knowledge do we hope to glean from running an unsupervised learning model on this dataset?
- What data is available? What type? What is missing? What can be removed?
- Is the data in a format that can be passed into an unsupervised learning model?
- Can I quickly hand off this data for others to use?
- What is the best way to cluster the data?
- What is the best way to visualize my results.
In order to complete this project, I am performing unsupervised machine learning functions on data provided CryptoCompare.
3D scatter plot using Plotly Express to plot the clusters. We were able to identify 4 Clusters as determined by the elbow curve.