Mastering Machine Learning Algorithms - Second Edition

13

Component Analysis and Dimensionality Reduction

In this chapter, we're going to introduce the most common and important techniques to perform component analysis and dimensionality reduction. When working with large datasets, it's often necessary to optimize the performance of the algorithms, and one of the most reasonable ways of achieving this goal is to remove those features whose information content is negligible. The models discussed in this chapter allow us to perform a complete analysis of the components of a dataset and to select only those components that make a valuable contribution to the results. In particular, we're going to discuss the following topics:

• Factor analysis
• Principal Component Analysis (PCA), Kernel PCA, and Sparse ...