Inexact regularized Newton method for unconstrained optimization with rapid rate of convergence

• Tác giả: Van Vu Nguyen, Ngoc Nguyen Tran
• Ngôn ngữ: eng
• Thông tin xuất bản: Tạp chí Khoa học Trường Đại học Quy Nhơn, 2022
• Mô tả vật lý: tr.11
• Bộ sưu tập: Metadata
• ID: 329949

Tóm tắt

This paper proposes an inexact regularized Newton method for solving unconstrained optimizationproblems. The proposed algorithm belongs to the class of outer-inner iteration scheme. Instead of solvingexactly linear systems, iterative linear solver will be applied to find approximate search directions. We willshow that the inexact algorithm preserved the fast local convergence property of exact algorithms. Somenumerical experiments are also conducted to show the benefits of our proposed algorithm.