Latent factor analysis

Latent factor analysis (LFA) is another technique that helps you reduce the dimensionality of the dataset. The overall idea is similar to PCA. However, in this case, there's no orthogonal decomposition of the input signal, and therefore, no output basis. Some data scientists think that LFA is a generalization of PCA that removes the constraint of orthogonality. Generally, LFA is used when a latent factor or a construct is expected to be present in the system. Under such a hypothesis, all of the features are observations of variables that are derived or influenced by the latent factor that is transformed linearly and which has an arbitrary waveform generator (AWG) noise. It is generally assumed that the latent factor ...

Get Python Data Science Essentials - Third Edition now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.