A fast al/lorithm for solving large scale MV (mean-variance) portfolio optimization problems is proposed. It is shown that by using T independent dat.a representing the rate of return of the assets, the MV model consisting of n assets can be put into a quadratic program with n + T variables, T linear const.raints and T quadratic terms in the objective function. As a result, the computation t.im...

This paper addresses the problem of curtailing the number of control actions and minimizing controller movements for real-time voltagelreactive power control. Algorithms are proposed to identify the most effective subset of control actions and to minimize controller movements. An algorithmic objective function appropriate for the treatment of system security and economy is also proposed. It eli...

This paper examines an integrated ratemaking scheme including a priori risk classification and a posteriori experience rating. In order to avoid the high penalties implied by the quadratic loss function, the symmetry between the overcharges and the undercharges is broken by introducing parametric loss functions of exponential type.

The dual formulation of the support vector machine (SVM) objective function is an instance of a nonnegative quadratic programming problem. We reformulate the SVM objective function as a matrix factorization problem which establishes a connection with the regularized nonnegative matrix factorization (NMF) problem. This allows us to derive a novel multiplicative algorithm for solving hard and sof...

In this paper we consider a guaranteed cost control problem for a class of uncertain discrete-time systems which contain structured uncertainties. The uncertainties are assumed to satisfy a certain sum quadratic constraint. The controller will be a minimax controller in the sense that it minimizes the maximum value of a cost function. For a given initial condition, the minimax optimal controlle...

where μ(Pi,Pj) = ∑ ck∈Pi,cl∈Pj σ(ck, cl). We now transform this simplified VCtoS problem into the quadratic assignment (QAP) problem, which is known to be NP-hard [SG76], as follows. Given n facilities denoted by set NF , n locations denoted by set NL, a flow matrix A, where each entry Aij represents the flow of materials moving from facility i to facility j, and a distance matrix D, where each...

This paper presents an improved lower bound and an approximation algorithm based on spectral decomposition for the binary constrained quadratic programming problem. To decompose spectrally the quadratic matrix in the objective function, we construct a low rank problem that provides a lower bound. Then an approximation algorithm for the binary quadratic programming problem together with a worst ...

Pareto Local Search (PLS) is a basic building block in many multiobjective metaheuristics. In this paper, Parallel Pareto Local Search based on Decomposition (PPLS/D) is proposed. PPLS/D decomposes the original search space into L subregions and executes L parallel search processes in these subregions simultaneously. Inside each subregion, the PPLS/D process is first guided by a scalar objectiv...

An instance of the quadratic assignment problem (QAP) with cost matrix Q is said to be linearizable if there exists an instance of the linear assignment problem (LAP) with cost matrix C such that for each assignment, the QAP and LAP objective function values are identical. The QAP linearization problem can be solved in O(n) time. However, for the special cases of Koopmans-Beckmann QAP and the m...

We propose an efficient heuristic for the constraint-driven communication synthesis (CDCS) of on-chip communication networks. The complexity of the synthesis problems comes from the number of constraints that have to be considered. In this paper we propose to cluster constraints to reduce the number that needs to be considered by the optimization algorithm. Then a quadratic programming approach...