Skip to content

Latest commit

 

History

History
31 lines (20 loc) · 2.03 KB

README.md

File metadata and controls

31 lines (20 loc) · 2.03 KB

L1-Norm-Algorithms

The current repository provides the code for the popular L1-norm Pricipal Component Analysis for matrix and tensor data sets.

The matrix algorithm was developed and published by P. P. Markopoulos, S. Kundu, S. Chamadia and D. A. Pados, "Efficient L1-Norm Principal-Component Analysis via Bit Flipping", in IEEE Transactions on Signal Processing, vol. 65, no. 16, pp. 4252-4264, Aug. 2017.

The tensor algorithm was developed and published by K. Tountas, D. A. Pados, M. J. Medley, "Conformity Evaluation and L1-norm Principal-Component Analysis of Tensor Data", in SPIE Big Data: Learning, Analytics, and Applications Conf., SPIE Defence and Commercial Sensing, Baltimore, MD, Mar. 2019.

Python Implementation:

The entry point for the matrix decomposition algorithm is the file l1_pca_example.py. The entry point for the tensor decomposition algorithm is the file ir_tensor_l1pca_example.py.

We have tested the code on Python 3.7.*. The prerequisite packages to run it are:

The prerequisite packages can be installed via pip: pip install -r requirements.txt

List of files included: l1_pca_example.py, l1pca_sbfk_v0.py, ir_tensor_l1pca_v0.py, ir_tensor_l1pca_example.py

MATLAB Implementation:

The entry point for the matrix decomposition algorithm is the file l1_pca_example.m. The entry point for the tensor decomposition algorithm is the file ir_tensor_l1pca_example.m.

We have tested the code on MatlabR2019a. The prerequisite packages to run it are:

List of files included: l1_pca_example.m, l1pca_BF.m, ir_tensor_l1pca_stable.m, ir_tensor_l1pca_example.m