Handling correlated and repeated measurements with the smoothed multivariate square-root Lasso Quentin Bertrand

The resulting optimization problem is convex and amenable, thanks to smoothing theory, to state-of-the-art optimization techniques that leverage the sparsity of the solutions.