In this lesson, you'll get a brief recap of key concepts from this section.

You will be able to:

• Highlight key takeaways concerning PCA

• Recall that PCA is a dimensionality reduction technique
• There are four steps to calculating PCA:
  • Center each feature by subtracting the feature mean
  • Calculate the covariance matrix for your normalized dataset
  • Calculate the eigenvectors/eigenvalues for the covariance matrix
    • Reorder your eigenvectors based on their accompanying eigenvalues (descending order to the eigenvalues)
  • Take the dot product of the transpose of the eigenvectors with the transpose of the normalized data
• You can also easily implement PCA using sci-kit learn

PCA is an essential tool in your skill set. Use it when you need to compress data, and experiment with its effect on machine learning algorithms as a preprocessing step. With PCA, you've now been exposed to unsupervised learning techniques which do not require labeled data.