TY - JOUR
T1 - A feasible interior-point method with full-Newton step for P*(κ)-weighted linear complementarity problem via the algebraically equivalent transformation
AU - Chi, Xiaoni
AU - Gan, Lin
AU - Gao, Zhuoran
AU - Chen, Jein Shan
N1 - Publisher Copyright:
© 2025 IMACS.
PY - 2026/2
Y1 - 2026/2
N2 - This paper investigates feasible interior-point method (IPM) with full-Newton step for P*(κ)-weighted linear complementarity problem (WLCP). In particular, by applying the algebraically equivalent transformation (AET) for linear optimization, we obtain the new search directions by solving the perturbed Newton system. The AET of the Newton system is based on the kernel function φ(t)=t−t, which is used for solving WLCP for the first time. At each iteration, our algorithm takes only full-Newton steps. Therefore, no line-searches are needed to update the iterates. We show the strict feasibility of the full-Newton step and the polynomial iteration complexity of our algorithm under suitable assumptions. Some numerical experiments demonstrate the effectiveness of the proposed algorithm.
AB - This paper investigates feasible interior-point method (IPM) with full-Newton step for P*(κ)-weighted linear complementarity problem (WLCP). In particular, by applying the algebraically equivalent transformation (AET) for linear optimization, we obtain the new search directions by solving the perturbed Newton system. The AET of the Newton system is based on the kernel function φ(t)=t−t, which is used for solving WLCP for the first time. At each iteration, our algorithm takes only full-Newton steps. Therefore, no line-searches are needed to update the iterates. We show the strict feasibility of the full-Newton step and the polynomial iteration complexity of our algorithm under suitable assumptions. Some numerical experiments demonstrate the effectiveness of the proposed algorithm.
KW - Algebraically equivalent transformation
KW - Feasible interior-point method
KW - Full-Newton step
KW - P(κ)-weighted linear complementarity problem
KW - Polynomial complexity
UR - https://www.scopus.com/pages/publications/105020596251
UR - https://www.scopus.com/pages/publications/105020596251#tab=citedBy
U2 - 10.1016/j.apnum.2025.10.003
DO - 10.1016/j.apnum.2025.10.003
M3 - Article
AN - SCOPUS:105020596251
SN - 0168-9274
VL - 220
SP - 144
EP - 166
JO - Applied Numerical Mathematics
JF - Applied Numerical Mathematics
ER -