In SVD decomposition of a term-document matrix (A = UΣVᵀ), what do the singular values in Σ represent?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — b is correct because singular values represent the magnitude of each latent dimension — larger singular values indicate the corresponding topic dimension explains more of the overall variation in term-document co-occurrence patterns. A is wrong; singular values do not count documents.
Full explanation below image
Full Explanation
B is correct because singular values represent the magnitude of each latent dimension — larger singular values indicate the corresponding topic dimension explains more of the overall variation in term-document co-occurrence patterns. A is wrong; singular values do not count documents. C is wrong; average term frequency is not represented by singular values. D is wrong; IDF weights are applied before SVD, not represented as singular values.