Cryptocurrency Clusters

Machine_Learning_Unsupervised

Background

As part of the Advisory Services Team of a financial consultancy. One of the clients, a prominent investment bank, is interested in offering a new cryptocurrency investment portfolio for its customers. The company, however, is lost in the vast universe of cryptocurrencies. We are tasked to create a report that includes what cryptocurrencies are on the trading market and determine whether they can be grouped to create a classification system for this new investment.

Instructions: Data Preparation

• Read crypto_data.csv into Pandas. The dataset was obtained from CryptoCompare.

• Discard all cryptocurrencies not being traded by filtering for currencies that are currently being traded. Next, drop the IsTrading column from the data frame.

• Remove all rows that have at least one null value.

• Filter for cryptocurrencies that have been mined.

• Convert data to be comprehensible to a machine learning algorithm, change to numeric values, and delete unnecessary columns.

• Standardize the dataset so columns containing larger values do not unduly influence the outcome.

Dimensionality Reduction

• Creating dummy variables
• Perform dimensionality reduction with PCA. Rather than specify
• Using PCA(n_components=0.99) create a model that will preserve approximately 99% of the explained variance,
• Reduce the dataset dimensions with t-SNE and examine scatter plot for distinct clusters

Cluster Analysis with K-Means

• Create an elbow plot to identify the best number of clusters

Recommendation

We can observe the shape of the elbow curve at 4, according to the model our K value optimal number of clusters is 4.