1: \begin{abstract}%

2: The structure of many real-world optimization problems includes

3: minimization of a nonlinear (or quadratic) functional subject to

4: bound and singly linear constraints (in the form of either equality

5: or bilateral inequality) which are commonly called as continuous

6: knapsack problems. Since there are efficient methods to solve

7: large-scale bound constrained nonlinear programs, it is desirable to

8: adapt these methods to solve knapsack problems, while preserving

9: their efficiency and convergence theories. The goal of this paper is

10: to introduce a general framework to extend a box-constrained

11: optimization solver to solve knapsack problems. This framework

12: includes two main ingredients which are O(n) methods; in terms of

13: the computational cost and required memory; for the projection onto

14: the knapsack constrains and the null-space manipulation of the

15: related linear constraint. The main focus of this work is on the

16: extension of Hager-Zhang active set algorithm (SIAM J. Optim. 2006,

17: pp. 526--557). The main reasons for this choice was its promising

18: efficiency in practice as well as its excellent convergence theories

19: (e.g., superlinear local convergence rate without strict

20: complementarity assumption). Moreover, this method does not use any

21: explicit form of second order information and/or solution of linear

22: systems in the course of optimization which makes it an ideal choice

23: for large-scale problems. Moreover, application of Birgin and

24: Mart{\'\i}nez active set algorithm (Comp. Opt. Appl. 2002, pp.

25: 101--125) for knapsack problems is also briefly commented. The

26: feasibility of the presented algorithm is supported by numerical

27: results for topology optimization problems.

28: 

29: \end{abstract}%