unfolding-linear 0.1.0

Deep unfolding of iterative methods to solve linear equations

Deep unfolding of iterative method

The package includes iterative methods for solving linear equations. However, due to the various parameters and performance of the iterative approach, it is necessary to optimize these parameters to improve the convergence rate. Such a proposed tool called deep_unfolding, which takes an iterative algorithm with a fixed number of iterations T, unravels its structure and adds trainable parameters. These parameters are then trained using deep learning techniques such as loss functions, stochastic gradient descent, and back-propagation. The package contains two different Iterative methods. The first package is called methods, which contains the conventional iterative method. The other package is called train_methods, which contains the deep unfolding of the iterative method.

Installation

pip install --upgrade pip
pip install deep-unfolding

Quick start

from deep_unfolding.train_methods import SORNet 

model = SORNet()

The diagram of the Deep unfolded network (DUN)

The Rest of package

Reference

If you use this software, please cite the following reference:

License

GPL License