Spectral_Sparsity_Graph_Learning

Implementation of the 2 main algorithms in "Graph learning under Spectral Sparsity Constraints", B. Subbareddy, Aditya Siripuram, Jingxin Zhang.

Model of the Graph Spectral Sparsity

Given Graph signal sequence $X \in \mathbb{R}^{N \times M}$ on graph $\mathcal{G} = \mathcal{G}(\mathcal{V},\mathcal{E})$ with $|V| = N$, we consider the following:

$$ X = V_{\mathcal{G}}Y + \eta $$

where $Y$ is k-sparse and is the Graph Fourier Transform of $X$.

The following optimization problems are hence drafted