Hank is a Senior Data Scientist at Boeing and a long time user of the R language. Prior to his current role, he led the Customer Data Science team at H2O.ai, a leading provider of machine learning and predictive analytics services.
Many times in machine learning, the goal is to find patterns in data without trying to make predictions. This is called unsupervised learning. One common use case of unsupervised learning is grouping consumers based on demographics and purchasing history to deploy targeted marketing campaigns. Another example is wanting to describe the unmeasured factors that most influence crime differences between cities. This course provides a basic introduction to clustering and dimensionality reduction in R from a machine learning perspective, so that you can get from data to insights as quickly as possible.
The k-means algorithm is one common approach to clustering. Learn how the algorithm works under the hood, implement k-means clustering in R, visualize and interpret the results, and select the number of clusters when it's not known ahead of time. By the end of the chapter, you'll have applied k-means clustering to a fun "real-world" dataset!
Hierarchical clustering is another popular method for clustering. The goal of this chapter is to go over how it works, how to use it, and how it compares to k-means clustering.
Principal component analysis, or PCA, is a common approach to dimensionality reduction. Learn exactly what PCA does, visualize the results of PCA with biplots and scree plots, and deal with practical issues such as centering and scaling the data before performing PCA.
The goal of this chapter is to guide you through a complete analysis using the unsupervised learning techniques covered in the first three chapters. You'll extend what you've learned by combining PCA as a preprocessing step to clustering using data that consist of measurements of cell nuclei of human breast masses.