Two-group classification via a biobjective margin maximization model

In this paper we propose a biobjective model for two-group classification via margin maximization, in which the margins in both classes are simultaneously maximized. The set of Pareto-optimal solutions is described, yielding a set of parallel hyperplanes, one of which is just the solution of the classical SVM approach.In order to take into account different misclassification costs or a priori probabilities, the ROC curve can be used to select one out of such hyperplanes by expressing the adequate tradeoff for sensitivity and specificity. Our result gives a theoretical motivation for using the ROC approach in case misclassification costs in the two groups are not necessarily equal.     

Quasiconcave vector maximization: Connectedness of the sets of Pareto-optimal and weak Pareto-optimal alternatives

We study the sets of Pareto-optimal and weakly Pareto-optimal solutions to a vector maximization problem defined by a continuous vector-valued quasiconcave criterion functionf and a closed convex set of alternativesS. IfS is compact, it is shown that the set of weakly Pareto-optimal alternatives is connected, but that the set of Pareto-optimal alternatives is not necessarily connected. However, the set of Pareto optima is shown to be connected for some important subclasses of quasiconcave functions. We also provide some reasonable conditions under which the compactness assumption onS may be relaxed and connectedness maintained.     

How relevant is the lack of reciprocity in pairwise comparisons? An experiment with AHP

We consider the 1-median problem with euclidean distances with uncertainty in the weights, expressed as possible changes within given bounds and a single budget constraint on the total cost of change. The upgrading (resp. downgrading) problem consists of minimizing (resp. maximizing) the optimal 1-median objective value over these weight changes. The upgrading problem is shown to belong to the family of continuous single facility location-allocation problems, whereas the downgrading problem reduces to a convex but highly non-differentiable optimization problem. Several structural properties of the optimal solution are proven for both problems, using novel planar partitions, the knapsack Voronoi diagrams, and lead to polynomial time solution algorithms.     

Multi-group support vector machines with measurement costs: A biobjective approach

Support Vector Machine has shown to have good performance in many practical classification settings. In this paper we propose, for multi-group classification, a biobjective optimization model in which we consider not only the generalization ability (modeled through the margin maximization), but also costs associated with the features. This cost is not limited to an economical payment, but can also refer to risk, computational effort, space requirements, etc. We introduce a Biobjective Mixed Integer Problem, for which Pareto optimal solutions are obtained. Those Pareto optimal solutions correspond to different classification rules, among which the user would choose the one yielding the most appropriate compromise between the cost and the expected misclassification rate.     

Optimization over the efficient set

This paper deals with the problem of maximizing a function over the efficient set of a linear multiple objective program. The approach is to formulate a biobjective program with an appropriate efficient set. The penalty function approach is motivated by an auxiliary problem due to Benson.     

Optimal risk sharing with general deviation measures

An optimal risk sharing problem for agents with utility functionals depending only on the expected value and a deviation measure of an uncertain payoff has been studied. The agents are assumed to have no initial endowments. A set of Pareto-optimal solutions to the problem has been characterized, and a particular solution from the set has been suggested. If an equilibrium exists, the suggested solution coincides with an equilibrium solution. As special cases, the optimal risk sharing problem in the form of expected gain maximization and the problem with a linear mean-deviation utility functional including averse and coherent risk measures have been addressed. In the case of expected gain maximization, the existence of an equilibrium has been shown.     

Control-space properties of cooperative games

If two or more players agree to cooperate while playing a game, they help one another to minimize their respective costs as long as it is not to their individual disadvantages. This leads at once to the concept of undominated solutions to a game. Anundominated orPareto-optimal solution has the property that, compared to any other solution, at least one playerdoes worse or alldo the same if they use a solution other than the Pareto-optimal one.Closely related to the concept of a Pareto-optimal solution is theabsolutely cooperative solution. Such a solution has the property that, compared to any other permissible solution,every playerdoes no better if a solution other than the absolutely cooperative one is employed.This paper deals with control-space properties of Pareto-optimal and absolutely cooperative solutions for both static, continuous games and differential games. Conditions are given for cases in which solutions to the Pareto-optimal and absolutely cooperative games lie in the interior or on the boundary of the control set.The solution of a Pareto-optimal or absolutely cooperative game is related to the solution of a minimization problem with avector cost criterion. The question of whether or not a problem with a vector cost criterion can be reduced to a family of minimization problems with ascalar cost criterion is also discussed.An example is given to illustrate the theory.This research was supported in part by NASA Grant No. NGR-03-002-011 and ONR Contract No. N00014-69-A-0200-1020.     

Duality for Anticonvex Programs

Calling anticonvex a program which is either a maximization of a convex function on a convex set or a minimization of a convex function on the set of points outside a convex subset, we introduce several dual problems related to each of these problems. We give conditions ensuring there is no duality gap. We show how solutions to the dual problems can serve to locate solutions of the primal problem.     

Nonlinear bipartite matching

We study the problem of optimizing nonlinear objective functions over bipartite matchings. While the problem is generally intractable, we provide several efficient algorithms for it, including a deterministic algorithm for maximizing convex objectives, approximative algorithms for norm minimization and maximization, and a randomized algorithm for optimizing arbitrary objectives.     

An approximation theory of matrix rank minimization and its application to quadratic equations

Matrix rank minimization problems are gaining plenty of recent attention in both mathematical and engineering fields. This class of problems, arising in various and across-discipline applications, is known to be NP-hard in general. In this paper, we aim at providing an approximation theory for the rank minimization problem, and prove that a rank minimization problem can be approximated to any level of accuracy via continuous optimization (especially, linear and nonlinear semidefinite programming) problems. One of the main results in this paper shows that if the feasible set of the problem has a minimum rank element with the least Frobenius norm, then any accumulation point of solutions to the approximation problem, as the approximation parameter tends to zero, is a minimum rank solution of the original problem. The tractability under certain conditions and convex relaxation of the approximation problem are also discussed. An immediate application of this theory to the system of quadratic equations is presented in this paper. It turns out that the condition for such a system without a nonzero solution can be characterized by a rank minimization problem, and thus the proposed approximation theory can be used to establish some sufficient conditions for the system to possess only zero solution.     

A D.C. biobjective location model

In this paper we address the biobjective problem of locating a semiobnoxious facility, that must provide service to a given set of demand points and, at the same time, has some negative effect on given regions in the plane. In the model considered, the location of the new facility is selected in such a way that it gives answer to these contradicting aims: minimize the service cost (given by a quite general function of the distances to the demand points) and maximize the distance to the nearest affected region, in order to reduce the negative impact. Instead of addressing the problem following the traditional trend in the literature (i.e., by aggregation of the two objectives into a single one), we will focus our attention in the construction of a finite -dominating set, that is, a finite feasible subset that approximates the Pareto-optimal outcome for the biobjective problem. This approach involves the resolution of univariate d.c. optimization problems, for each of which we show that a d.c. decomposition of its objective can be obtained, allowing us to use standard d.c. optimization techniques.     

Target Level Method in Linear Multi-Criterion Problems

We consider the target level method for solving linear multi-criterion maximization problems. The method finds an efficient (Pareto-optimal) vector estimate that is closest in the Chebyshev metric to the target level point specified by the decision maker. The proposed method describes (parametrizes and approximates) the efficient set. In the linear case the number of scalar optimization problems needed to describe the set of efficient vector estimates is substantially reduced. A formula is derived which, under certain conditions, can be used to compute efficient vector estimates without solving any optimization problems. An algorithm based on these results is proposed for two-criterion problems.     

Multisourcing suppliers selection in service outsourcing

Multisourcing suppliers selection in service outsourcing involves selecting a supplier portfolio with a reasonable number of suppliers and better performance to cover aspiration levels of criteria. It is a specific weighted matching problem with new challenges. This paper proposes a decision method for solving this problem. In the proposed method, different formats of preference information, including numerical values, interval numbers and linguistic variables, are used to express alternative ratings. The technique for order preference by similarity to ideal solution is extended to aggregate the three formats of preference information. A bi-objective 0–1 linear programming model using the aggregated information is built to select a desired supplier portfolio, in which the objectives of minimization of suppliers number and maximization of supplier performance are involved. To solve this model, we transform it into an equivalent, and then an exact multi-objective branch-and-bound algorithm is developed to obtain Pareto-optimal solutions. In addition, a real case of an insurance company is used to illustrate the applicability of the proposed method.     

Dominators for Multiple-objective Quasiconvex Maximization Problems

In this paper we address the problem of finding a dominator for a multiple-objective maximization problem with quasiconvex functions. The one-dimensional case is discussed in some detail, showing how a Branch-and-Bound procedure leads to a dominator with certain minimality properties. Then, the well-known result stating that the set of vertices of a polytope S contains an optimal solution for single-objective quasiconvex maximization problems is extended to multiple-objective problems, showing that, under upper-semicontinuity assumptions, the set of (k 21)-dimensional faces is a dominator for k-objective problems. In particular, for biobjective quasiconvex problems on a polytope S, the edges of S constitute a dominator, from which a dominator with minimality properties can be extracted by Branch-and Bound methods.     

Optimization of chance constraint programming with sum-of-fractional objectives – An application to assembled printed circuit board problem

This paper presents a procedure to solve a chance constraint programming problem with sum-of-fractional objectives. The problem and the solution procedure are described with the help of a practical problem – assembled printed circuit boards (PCBs). Errors occurring during assembling PCBs are in general classified into three kinds, viz. machine errors, manual errors and other errors. These errors may lead to the rejection of the major portion of the production and hence result the manufacturer a huge loss. The problem is decomposed to have two objective functions; one is a sum-of-fractional objectives and the other is a non-linear objective with bounded constraints. The interest is to maximize the sum-of-fractional objectives and to minimize the non-linear objective, subject to stochastic and non-stochastic bounded environment. The first problem deals with the maximization of the profit (maximizing sum-of-fractional objectives) and the second deals with the minimization of the loss (errors), so as to obtain the maximum profit after removing the number of defective PCBs.     

A Semidefinite Programming approach for solving Multiobjective Linear Programming

Several algorithms are available in the literature for finding the entire set of Pareto-optimal solutions of Multiobjective Linear Programmes (MOLPs). However, all of them are based on active-set methods (simplex-like approaches). We present a different method, based on a transformation of any MOLP into a unique lifted Semidefinite Program (SDP), the solutions of which encode the entire set of Pareto-optimal extreme point solutions of any MOLP. This SDP problem can be solved, among other algorithms, by interior point methods; thus unlike an active set-method, our method provides a new approach to find the set of Pareto-optimal solutions of MOLP.     

Maximizing breaks and bounding solutions to the mirrored traveling tournament problem

We investigate the relation between two aspects of round robin tournament scheduling problems: breaks and distances. The distance minimization problem and the breaks maximization problem are equivalent when the distance between every pair of teams is equal to 1. We show how to construct schedules with a maximum number of breaks for some tournament types. The connection between breaks maximization and distance minimization is used to derive lower bounds to the mirrored traveling tournament problem and to prove the optimality of solutions found by a heuristic for the latter.     

Unequal probability sampling from a finite population: A multicriteria approach

In this note we address the problem of determining selection probabilities for multipurpose surveys, when the aim is the simultaneous minimization of variances for each variable under study. A characterization of the set of Pareto-optimal designs is given for designs with replacement and also for a class of designs without replacement, namely, Poisson designs.     

Performance guarantees of greedy algorithms for the problems on hereditary systems

Under study are the problems of maximization and minimization of additive functions on hereditary systems which generalize many computationally hard combinatorial optimization problems. A performance guarantee of the greedy algorithm is proven in terms of the parameters of a feasible set and the objective function of the maximization problem. This bound improves the well-known Jenkyns—Korte—Hausmann bound. An analogous result is obtained for the minimization problem of an additive function on a hereditary system.     

Null space conditions and thresholds for rank minimization

Minimizing the rank of a matrix subject to constraints is a challenging problem that arises in many applications in machine learning, control theory, and discrete geometry. This class of optimization problems, known as rank minimization, is NP-hard, and for most practical problems there are no efficient algorithms that yield exact solutions. A popular heuristic replaces the rank function with the nuclear norm—equal to the sum of the singular values—of the decision variable and has been shown to provide the optimal low rank solution in a variety of scenarios. In this paper, we assess the practical performance of this heuristic for finding the minimum rank matrix subject to linear equality constraints. We characterize properties of the null space of the linear operator defining the constraint set that are necessary and sufficient for the heuristic to succeed. We then analyze linear constraints sampled uniformly at random, and obtain dimension-free bounds under which our null space properties hold almost surely as the matrix dimensions tend to infinity. Finally, we provide empirical evidence that these probabilistic bounds provide accurate predictions of the heuristic’s performance in non-asymptotic scenarios.