I believe while you are right in general, she is looking at a class of problems for which tensors handily triumph other methods. You might be interested in thes

I believe while you are right in general, she is looking at a class of problems for which tensors handily triumph other methods. You might be interested in these papers - http://newport.eecs.uci.edu/anandkumar/pubs/powerdynamics.pdf http://newport.eecs.uci.edu/anandkumar/pubs/powerdynamics.pd... http://newport.eecs.uci.edu/anandkumar/pubs/ProvableNN_sparse.pdf http://newport.eecs.uci.edu/anandkumar/pubs/ProvableNN_spars...