Components Weights is a term used in the text analytics industry to refer to the singular values of an LSA model. The component weights of an LSA model are the square of the singular values. ComponentWeights is a 1-by-NumComponents vector where the jth entry corresponds to the weight of component j. The components are ranked in decreasing weights.
Importance of Components Weights
The component weights of an LSA model are important because they can be used to estimate the importance of components. The weights can also be used to determine which components are most important for a particular application.
Components Weights vs. Other Similar Terms
Latent semantic analysis (LSA) weights and singular value decomposition (SVD) weights are other terms that may be used to refer to component weights. However, LSA weights and SVD weights are not the same as component weights. LSA weights are the singular values of an LSA model, while SVD weights are the singular values of an SVD decomposition. Component weights are the square of the singular values.