Efficient algorithms for fitting weighted least squares regression with \eqn{L_{1}}{L1} regularization on both the coefficients and weight vectors, which is able to perform simultaneous variable selection and outliers detection efficiently.
