Azam Kheyri - New Sparse Estimator for High-Dimensional Precision Matrix Estimation скачать в хорошем качестве

Azam Kheyri - New Sparse Estimator for High-Dimensional Precision Matrix Estimation

In recent years, there has been significant research into the problem of estimating covariance and precision matrices in high-dimensional settings. One approach to improving the accuracy of maximum likeli-hood estimation is through penalization of the log likeli-hood function. We propose a new estimator for preci-sion matrix estimation that uses an elastic net-type pen-alty and takes the target matrix into consideration. Our estimator combines the ridge and graphical lasso meth-ods, which allows for a more efficient estimation pro-cess. Numerical results demonstrate that our proposed method outperforms existing methods in terms of both estimation accuracy and computational efficiency.