On interval-subgradient and no-good cuts

Interval-gradient cuts are (nonlinear) valid inequalities derived from continuously differentiable nonconvex constraints. In this paper we define interval-subgradient cuts, a generalization to nondifferentiable constraints, and show that no-good cuts with 1-norm are a special case of interval-subgradient cuts. We then briefly discuss what happens if other norms are used.     

On mixed-integer sets with two integer variables

We show that every facet-defining inequality of the convex hull of a mixed-integer polyhedral set with two integer variables is a crooked cross cut (which we defined in 2010). We extend this result to show that crooked cross cuts give the convex hull of mixed-integer sets with more integer variables if the coefficients of the integer variables form a matrix of rank 2. We also present an alternative characterization of the crooked cross cut closure of mixed-integer sets similar to the one on the equivalence of different definitions of split cuts presented in Cook et al. (1990) [4]. This characterization implies that crooked cross cuts dominate the 2-branch split cuts defined by Li and Richard (2008) [8]. Finally, we extend our results to mixed-integer sets that are defined as the set of points (with some components being integral) inside a closed, bounded and convex set.     

Mixed-Integer Cuts from Cyclic Groups

We analyze a separation procedure for Mixed-Integer Programs related to the work of Gomory and Johnson on interpolated subadditive functions. This approach has its roots in the Gomory-Johnson characterization on the master cyclic group polyhedron. To our knowledge, the practical benefit that can be obtained by embedding interpolated subadditive cuts in a cutting plane algorithm was not investigated computationally by previous authors. In this paper we compute, for the first time, the lower bound value obtained when adding (implicitly) all the interpolated subadditive cuts that can be derived from the individual rows of an optimal LP tableau, thus approximating the optimization over the intersection of the Gomory corner polyhedron with the LP relaxation of the original problem formulation. The computed bound is compared with that obtained when only Gomory mixed-integer cuts are used, on a very large test-bed of MIP instances.     

Lift-and-project for general two-term disjunctions

In this paper we generalize the cut strengthening method of Balas and Perregaard for 0/1 mixed-integer programming to disjunctive programs with general two-term disjunctions. We apply our results to linear programs with complementarity constraints.     

Generating Cuts from Surrogate Constraint Analysis for Zero-One and Multiple Choice Programming

This paper presents a new surrogate constraint analysis that givesrise to a family of strong valid inequalities calledsurrogate-knapsack (S-K) cuts. The analytical procedure presentedprovides a strong S-K cut subject to constraining the values ofselected cut coefficients, including the right-hand side. Ourapproach is applicable to both zero-one integer problems and problemshaving multiple choice (generalized upper bound) constraints. We alsodevelop a strengthening process that further tightens the S-K cutobtained via the surrogate analysis. Building on this, we develop apolynomial-time separation procedure that successfully generates anS-K cut that renders a given non-integer extreme point infeasible. Weshow how sequential lifting processes can be viewed in our framework,and demonstrate that our approach can obtain facets that are notavailable to standard lifting methods. We also provide a relatedanalysis for generating fast cuts. Finally, we presentcomputational results of the new S-K cuts for solving 0-1 integerprogramming problems. Our outcomes disclose that the new cuts arecapable of reducing the duality gap between optimal continuous andinteger feasible solutions more effectively than standard liftedcover inequalities, as used in modern codes such as the CPLEX mixed0-1 integer programming solver.     

A note on the MIR closure and basic relaxations of polyhedra

Anderson et al. (2005) [1] show that for a polyhedral mixed integer set defined by a constraint system Ax≥b, along with integrality restrictions on some of the variables, any split cut is in fact a split cut for a basic relaxation, i.e., one defined by a subset of linearly independent constraints. This result implies that any split cut can be obtained as an intersection cut. Equivalence between split cuts obtained from simple disjunctions of the form x_j≤0 or x_j≥1 and intersection cuts was shown earlier for 0/1-mixed integer sets by Balas and Perregaard (2002) [4]. We give a short proof of the result of Anderson, Cornuéjols and Li using the equivalence between mixed integer rounding (MIR) cuts and split cuts.     

Number of minimum vertex cuts in transitive graphs

Let G be a k-regular vertex transitive graph with connectivity κ(G)=k and let m_k(G) be the number of vertex cuts with k vertices. Define m(n,k)=min{m_k(G): GT_n,k}, where T_n,k denotes the set of all k-regular vertex transitive graphs on n vertices with κ(G)=k. In this paper, we determine the exact values of m(n,k).     

The strength of multi-row aggregation cuts for sign-pattern integer programs

Sign-pattern IPs are a generalization of packing IPs, where for a given column all coefficients are either non-negative or non-positive. We show that the aggregation closure for such sign-pattern IPs can be 2-approximated by the original 1-row closure. This generalizes a result for packing IPs. On the other hand, unlike in the case of packing IPs, we show that the multi-row aggregation closure cannot be well approximated by the original multi-row closure.     

A note on packing paths in planar graphs

Seymour (1981) proved that the cut criterion is necessary and sufficient for the solvability of the edge-disjoint paths problem when the union of the supply graph and the demand graph is planar and Eulerian. When only planarity is required, Middendorf and Pfeiffer (1993) proved the problem to be NP-complete. For this case, Korach and Penn (1992) proved that the cut criterion is sufficient for the existence of a near-complete packing of paths. Here we generalize this result by showing how a natural strengthening of the cut criterion yields better packings of paths.Analogously to Seymour's approach, we actually prove a theorem on packing cuts in an arbitrary graph and then the planar edge-disjoint paths case is obtained by planar dualization. The main result is derived from a theorem of Seb (1990) on the structure of ±1 weightings of a bipartite graph with no negative circuits.Research partially supported by the Hungarian National Foundation for Scientific Research Grants OTKA 2118 and 4271.     

An O(K·n4) algorithm for finding the k best cuts in a network

An algorithm for finding the K best cuts in a network is presented. Using a branch technique introduced by Lawler [4] we reduce the problem to K computations of 2nd best cuts. The latter problem can be solved by an O(n⁴) algorithm yielding an overall complexity of O(K·n⁴) for the presented algorithm.     

Adaptive multicut aggregation for two-stage stochastic linear programs with recourse

Outer linearization methods for two-stage stochastic linear programs with recourse, such as the L-shaped algorithm, generally apply a single optimality cut on the nonlinear objective at each major iteration, while the multicut version of the algorithm allows for several cuts to be placed at once. In general, the L-shaped algorithm tends to have more major iterations than the multicut algorithm. However, the trade-offs in terms of computational time are problem dependent. This paper investigates the computational trade-offs of adjusting the level of optimality cut aggregation from single cut to pure multicut. Specifically, an adaptive multicut algorithm that dynamically adjusts the aggregation level of the optimality cuts in the master program, is presented and tested on standard large-scale instances from the literature. Computational results reveal that a cut aggregation level that is between the single cut and the multicut can result in substantial computational savings over the single cut method.     

Depth-Optimized Convexity Cuts

This paper presents a general, self-contained treatment of convexity or intersection cuts. It describes two equivalent ways of generating a cut—via a convex set or a concave function—and a partial-order notion of cut strength. We then characterize the structure of the sets and functions that generate cuts that are strongest with respect to the partial order. Next, we specialize this analytical framework to the case of mixed-integer linear programming (MIP). For this case, we formulate two kinds of the deepest cut generation problem, via sets or via functions, and subsequently consider some special cases which are amenable to efficient computation. We conclude with computational tests of one of these procedures on a large set of MIPLIB problems.     

The split closure of a strictly convex body

The Chvátal-Gomory closure and the split closure of a rational polyhedron are rational polyhedra. It has been recently shown that the Chvátal-Gomory closure of a strictly convex body is also a rational polytope. In this note, we show that the split closure of a strictly convex body is defined by a finite number of split disjunctions, but is not necessarily polyhedral. We also give a closed form expression in the original variable space of a split cut for full-dimensional ellipsoids.     

Disjunctive cuts for continuous linear bilevel programming

This work shows how disjunctive cuts can be generated for a bilevel linear programming problem (BLP) with continuous variables. First, a brief summary on disjunctive programming and bilevel programming is presented. Then duality theory is used to reformulate BLP as a disjunctive program and, from there, disjunctive programming results are applied to derive valid cuts. These cuts tighten the domain of the linear relaxation of BLP. An example is given to illustrate this idea, and a discussion follows on how these cuts may be incorporated in an algorithm for solving BLP.     

Intersection properties of maximal directed cuts in digraphs

If $D$ is a finite digraph, a directed cut is a subset of arcs in $D$ having tail in some subset $X?V\left(D\right)$ and head in $V\left(D\right)?X$. In this paper we prove two general results concerning intersections between maximal paths, cycles and maximal directed cuts in $D$. As a direct consequence of these results, we deduce that there is a path, or a cycle, in $D$ that crosses each maximal directed cut.