Principal Component Analysis (PCA) is the most widely used linear method in dealing with dimension reduction problems. It is useful when data contains many features, and there is redundancy (correlation) within these features. To remove redundant features, PCA maps high dimension data into lower dimensions by reducing features into a smaller number of principal components that account for most of the variance of the original features. In this recipe, we will look at how to perform dimension reduction with the PCA method.
In this recipe, we will use the swiss
dataset as our target to perform PCA. The swiss
dataset includes standardized fertility measures and socio-economic indicators from around the year 1888 for each of the 47 French-speaking provinces of Switzerland.