Robust non-negative least squares using sparsity

摘要: Sparse, non-negative signals occur in many applications. To recover such signals, estimation posed as least squares problems have proven to be fruitful. Efficient algorithms with high accuracy been proposed, but of them assume either perfect knowledge the dictionary generating signal, or attempts explain deviations from this by attributing components that for some reason is missing dictionary. In work, we propose a robust algorithm allows differ assumed dictionary, introducing uncertainty setup. The proposed enables an improved modeling measurements, and may efficiently implemented using ADMM implementation. Numerical examples illustrate performance compared standard LASSO estimator.