Summary

This chapter introduced you to two widely used unsupervised, clustering algorithms, HCA and k-means clustering. While learning about k-means clustering, we leveraged the power of loops to create ensembles of models for tuning the number of clusters and to gain more confidence in our predictions. During the PCA section, we determined the number of principal components for dimensionality reduction and fit the components to a k-means model. Additionally, we compared the differences in k-means model performance before and after PCA transformation. We were introduced to an algorithm, LDA, which reduces dimensionality in a supervised manner. Lastly, we tuned the number of components in LDA by iterating through all possible values for components and programmatically returning the value resulting in the best accuracy score from a Random Forest classifier model. You should now feel comfortable with dimensionality reduction and unsupervised learning techniques.

We were briefly introduced to...