Stratified Non-Negative Tensor Factorization
This is an incremental improvement for researchers analyzing multi-modal data with stratified sources, enabling more efficient topic identification.
The paper tackled the problem of applying stratified non-negative factorization to multi-modal data without losing tensor structure, by extending it to tensors with a multiplicative update rule, resulting in Stratified-NTF that identifies interpretable topics with lower memory requirements than Stratified-NMF.
Non-negative matrix factorization (NMF) and non-negative tensor factorization (NTF) decompose non-negative high-dimensional data into non-negative low-rank components. NMF and NTF methods are popular for their intrinsic interpretability and effectiveness on large-scale data. Recent work developed Stratified-NMF, which applies NMF to regimes where data may come from different sources (strata) with different underlying distributions, and seeks to recover both strata-dependent information and global topics shared across strata. Applying Stratified-NMF to multi-modal data requires flattening across modes, and therefore loses geometric structure contained implicitly within the tensor. To address this problem, we extend Stratified-NMF to the tensor setting by developing a multiplicative update rule and demonstrating the method on text and image data. We find that Stratified-NTF can identify interpretable topics with lower memory requirements than Stratified-NMF. We also introduce a regularized version of the method and demonstrate its effects on image data.