An implementation of the subspace power method, in MATLAB and Python, for decomposing symmetric tensors, as described in
J. Kileel, J. M. Pereira, Subspace power method for symmetric tensor decomposition, arXiv:1912.04007
To compare performance with other tensor decomposition packages,
and reproduce the results obtained in the paper, the users will have to install
and download external packages themselves. An exception to this is the implementation
of the FOOBI algorithm, from Fourth-Order Cumulant-Based Blind Identification
of Underdetermined Mixtures,
for which we did not find a MATLAB implementation,
and implemented ourselves. If you need assistance setting this up, you can
send an e-mail to [email protected].
A README for the installation of other packages may also be added in the future.
The following are required packages:
-
GPCA-PDA (GPCA only)
-
GPCA-Voting (GPCA only)
To install you just need to add the folders
MATLAB\ and MATLAB\helper_functions\ to the MATLAB path. Alternatively, you may run the file MATLAB\setup.m.
- To reproduce the results in Subspace power method for symmetric tensor
decomposition, run the file
MATLAB\tests\run_SPM_paper.m.
numpyscipy
numba: This package provides a just-in-time pre-compiler that can considerably speed up SPM performance. It is used only if the numba package is installed.
To install you just need to copy the python files, and from SPM.py, import the method subspace_power_method.
Additional methods (such as generate_lowrank_tensor) are also available which can be useful for testing SPM.