Parallel Strategies for Solving SURE Models with Variance Inequalities and Positivity of Correlations Constraints

The problem of computing estimates of parameters in SURE models with variance inequalities and positivity of correlations constraints is considered. Efficient algorithms that exploit the block bi-diagonal structure of the data matrix are presented. The computational complexity of the main matrix factorizations is analyzed. A compact method to solve the model with proper subset regressors is proposed.