Implicit Regularization of Decentralized Gradient Descent for Decentralized Sparse Regression

Neural Information Processing Systems 

We consider learning a sparse model from linear measurements taken by a network of agents.