“…Recently, many researchers adopted ADMM algorithm to high-dimensional statistics and machine learning, see Yin et al [14], Xue et al [15], Zhang and Zou [16], Gu et al [17], and Kapla et al [18], and many other topics, such as matrix completion, tensor completion, and sparse recovery, see Li et al [19,20], Liu et al [21], and Shi et al [22]. Te main advantages of the ADMM algorithm are its fexibility at simplifying a diversity of optimization problems and its good convergence property, see Boyd et al [23].…”