This is the repository for the paper:
Neural Nonnegative CP Decomposition for Hierarchical Tensor Analysis
by Joshua Vendrow, Jamie Haddock, and Deanna Needell.Abstract: There is a significant demand for topic modeling on large-scale data with complex multi-modal structure in applications such as multi-layer network analysis, temporal document classification, and video data analysis; frequently this multi-modal data has latent hierarchical structure. We propose a new hierarchical nonnegative CANDECOMP/PARAFAC (CP) decomposition (hierarchical NCPD) model and a training method, Neural NCPD, for performing hierarchical topic modeling on multi-modal tensor data. Neural NCPD utilizes a neural network architecture and backpropagation to mitigate error propagation through hierarchical NCPD. Here, we present this hierarchical NCPD approach and demonstrate its efficacy on experiments using synthetic and temporal document data sets.