Neural Computation
October 2006, Vol. 18, No. 10, Pages 2293-2319
(doi: 10.1162/neco.2006.18.10.2293)
Images, Frames, and Connectionist Hierarchies
Article PDF (874.18 KB)
Abstract
The representation of hierarchically structured knowledge in systems using distributed patterns of activity is an abiding concern for the connectionist solution of cognitively rich problems. Here, we use statistical unsupervised learning to consider semantic aspects of structured knowledge representation. We meld unsupervised learning notions formulated for multilinear models with tensor product ideas for representing rich information. We apply the model to images of faces.