In this research, a modified hybrid Dai-Yuan and Hestenes-Stiefel (DYHS) conjugate gradient (CG) algorithm is presented, with develops sufficient descent property. The hyperplane projection method is also used and the algorithm is a convex combination of DY and HS parameters. With some reasonable assumptions, the global convergence property is obtained under accelerated step size. The numerical experiment demonstrate that the DYHS algorithm is more effective and robust than other existing methods for solving nonlinear monotone equations and signal recovery problems.
Mathematics Subject Classification: 65K05, 90C52, 90C26