new shrinkage estimation in t-regression model with stochasticsub-space restriction

This paper intruduces a new biased shrinkage estimator for estimation ofregression coeficients in the multivariate-t linear model with stochastic restrictions. itsbias and quadratic risk functions are derived approximately. it is evident from the numericalresults that the shrinkage estimator outperforms the generalized least squaresestimator in risk since.