Estimation of the Sample Covariance Matrix from Compressive Measurements

Paper: https://arxiv.org/abs/1512.08887

The program is an implementation of the above paper by Farhad Pourkamali-Anaraki (2016). The CovarianceMatrix class provides a biased and an unbiased estimator as proposed by the author and compares their accuracies in predicting the original covariance matrix.

Note: The input matrix X in the estimator is assumed to be a zero mean distribution.

For more details, please refer to the project proposal and report files attached in the repository.