Abstract:In this paper, we present a convergence analysis of the Group Projected Subspace Pursuit (GPSP) algorithm proposed by He et al. [HKL+23] (Group Projected subspace pursuit for IDENTification of variable coefficient differential equations (GP-IDENT), Journal of Computational Physics, 494, 112526) and extend its application to general tasks of block sparse signal recovery. We prove that when the sampling matrix satisfies the Block Restricted Isometry Property (BRIP) with a sufficiently small Block Restricted Isometry Constant (BRIC), GPSP exactly recovers the true block sparse signals. When the observations are noisy, this convergence property of GPSP remains valid if the magnitude of true signal is sufficiently large. GPSP selects the features by subspace projection criterion (SPC) for candidate inclusion and response magnitude criterion (RMC) for candidate exclusion. We compare these criteria with counterparts of other state-of-the-art greedy algorithms. Our theoretical analysis and numerical ablation studies reveal that SPC is critical to the superior performances of GPSP, and that RMC can enhance the robustness of feature identification when observations contain noises. We test and compare GPSP with other methods in diverse settings, including heterogeneous random block matrices, inexact observations, face recognition, and PDE identification. We find that GPSP outperforms the other algorithms in most cases for various levels of block sparsity and block sizes, justifying its effectiveness for general applications.

An efficient and robust varying-parameter projection neural network for sparse signal reconstruction

Approximate Message Passing with Nearest Neighbor Sparsity Pattern Learning

Convergence of Projected Subgradient Method with Sparse or Low-Rank Constraints

Time-varying neurodynamic optimization approaches with fixed-time convergence for sparse signal reconstruction

Matrix-Form Neural Networks for Complex-Variable Basis Pursuit Problem With Application to Sparse Signal Reconstruction

Variable projection algorithms with sparse constraint for separable nonlinear models

A distributed neurodynamic algorithm for sparse signal reconstruction via l 1 -minimization

Parameterized proximal-gradient algorithms for L1/L2 sparse signal recovery

Enhanced Expressive Power and Fast Training of Neural Networks by Random Projections

From Least Squares to Sparse: A Non-Convex Approach with Guarantee

Positive Competitive Networks for Sparse Reconstruction

A Unified Predefined-Time Convergent and Robust ZNN Model for Constrained Quadratic Programming

A Non-convex Approach for Sparse Recovery with Convergence Guarantee

Learning soft threshold for sparse reparameterization using gradual projection operators

Inverse-free distributed neurodynamic optimization algorithms for sparse reconstruction

Group Projected Subspace Pursuit for Block Sparse Signal Reconstruction: Convergence Analysis and Applications

Weakly Convex Regularized Robust Sparse Recovery Methods with Theoretical Guarantees

A Modified Spectral Conjugate Gradient Projection Method for Signal Recovery.

The Approximation of the Projection Pursuit Learning Networks for Non-linear Time Series

Dynamic Zero-Point Attracting Projection for Time-Varying Sparse Signal Recovery