Principal Component Analysis (PCA) is a statistical technique for dimensionality reduction that transforms data with high-dimensional space into low-dimensional space. Its goal is to replace a large number of correlated variables with a smaller number of uncorrelated variables while retaining as much information in the original variables as possible, hence playing a significant role in feature engineering tasks of the machine learning pipeline.
Principal Component Analysis
Getting ready
In this recipe, we will illustrate the technique of PCA using the USArrests dataset that contains crime-related statistics, such as Assault, Murder, Rape, and UrbanPop, per 100,000 residents in 50 states in the US.
If you have not already downloaded...