When we left the recommendation engine problem at the beginning of the chapter, we created the
wordsMatrix variable. We should now have enough background with regards to the benefits of the eigenvalue decomposition process, which will help us to finish the creation of our recommendation engine. We can use the
principalComponentAnalysis function and produce our own dataset based on
wordsMatrix, as follows:
> let pcaMatrix = principalComponentAnalysis wordsMatrix 5
pcaMatrix function is a compressed form of the
wordsMatrix variable, which focuses on the 5 vectors with the highest variance.
Each row in the
pcaMatrix function represents one user. If we were to plot this dataset (which ...