How Where and When we should use PCA

Perhaps the most popular dimensionality reduction technique in machine learning is Principal Component Analysis. Well, practically everything has already been written about the PCA. However, I have the impression that many articles on this topic are not complete, I always missed something in them. They are devoted to the mathematical side or conversely the implementation is presented. I am interested in using algorithms to add value to the business and to know how they work, to adapt them, and thus understand the data even better. If we can implement a machine learning algorithm in Python, R or SAS and if we know how it works then we can come up with new algorithms. Furthermore, there are Data Scientists who use this algorithm without basic knowledge of what mathematical computations occur, much less what the concept of how it works is.

If you working in Data Science you will have to start using PCA. Principal Component Analysis is beautiful because it is useful and easy to apply. In this article, I will want you to intuitively know what PCA is, how it works and why it is used. We will look at the mathematical as well as the practical side of how this machine learning algorithm works.