Separation oracle

A separation oracle (also called a cutting-plane oracle) is a concept in the mathematical theory of convex optimization. It is a method to describe a convex set that is given as an input to an optimization algorithm. Separation oracles are used as input to ellipsoid methods.: 87, 96, 98 

Definition

Let K be a convex and compact set in Rn. A strong separation oracle for K is an oracle (black box) that, given a vector y in Rn, returns one of the following:: 48 

A strong separation oracle is completely accurate, and thus may be hard to construct. For practical reasons, a weaker version is considered, which allows for small errors in the boundary of K and the inequalities. Given a small error tolerance d>0, we say that:

The weak version also considers rational numbers, which have a representation of finite length, rather than arbitrary real numbers. A weak separation oracle for K is an oracle that, given a vector y in Qn and a rational number d>0, returns one of the following:: 51 

Implementation

A special case of a convex set is a set represented by linear inequalities: K={x|Ax≤b}. Such a set is called a convex polytope. A strong separation oracle for a convex polytope can be implemented, but its run-time depends on the input format.

Representation by inequalities

If the matrix A and the vector b are given as input, so that K={x|Ax≤b}, then a strong separation oracle can be implemented as follows. Given a point y, compute Ay:

  • If the outcome is at most b, then y is in K by definition;
  • Otherwise, there is at least one row c of A, such that c⋅y is larger than the corresponding value in b; this row c gives us the separating hyperplane, as c⋅y>b≥c⋅x for all x in K.

This oracle runs in polynomial time as long as the number of constraints is polynomial.

Representation by vertices

Suppose the set of vertices of K is given as an input, so that K=conv(v1,…,vk)= the convex hull of its vertices. Then, deciding whether y is in K requires to check whether y is a convex combination of the input vectors, that is, whether there exist coefficients z1,...,zk such that: : 49 

  • z1⋅v1+⋯+zk⋅vk=y;
  • 0≤zi≤1 for all i in 1,...,k.

This is a linear program with k variables and n equality constraints (one for each element of y). If y is not in K, then the above program has no solution, and the separation oracle needs to find a vector c such that

  • c⋅y>c⋅vi for all i in 1,...,k.

Note that the two above representations can be very different in size: it is possible that a polytope can be represented by a small number of inequalities, but has exponentially many vertices (for example, an n-dimensional cube). Conversely, it is possible that a polytope has a small number of vertices, but requires exponentially many inequalities (for example, the convex hull of the 2n vectors of the form (0,...,±1,...,0).

Problem-specific representation

In some linear optimization problems, even though the number of constraints is exponential, one can still write a custom separation oracle that works in polynomial time. Some examples are:

  • The minimum-cost arborescence problem: given a weighted directed graph and a vertex r in it, find a subgraph of minimum cost that contains a directed path from r to any other vertex. The problem can be presented as an LP with a constraint for each subset of vertices, which is an exponential number of constraints. However, a separation oracle can be implemented using n-1 applications of the minimum cut procedure.
  • The maximum independent set problem. It can be approximated by an LP with a constraint for every odd-length cycle. While there are exponentially-many such cycles, a separation oracle that works in polynomial time can be implemented by just finding an odd cycle of minimum length, which can be done in polynomial time.
  • The dual of the configuration linear program for the bin packing problem. It can be approximated by an LP with a constraint for each feasible configuration. While there are exponentially-many such cycles, a separation oracle that works in pseudopolynomial time can be implemented by solving a knapsack problem. This is used by the Karmarkar-Karp bin packing algorithms.

Non-linear sets

Let f be a convex function on Rn. The set K={(x,t)|f(x)≤t} is a convex set in Rn+1. Given an evaluation oracle for f (a black box that returns the value of f for every given point), one can easily check whether a vector (y, t) is in K. In order to get a separation oracle, we need also an oracle to evaluate the subgradient of f.: 49  Suppose some vector (y, s) is not in K, so f(y) > s. Let g be the subgradient of f at y (g is a vector in Rn). Denote c:=(g,−1).Then, c⋅(y,s)=g⋅y−s>g⋅y−f(y), and for all (x, t) in K: c⋅(x,t)=g⋅x−t≤g⋅x−f(x). By definition of a subgradient: f(x)≥f(y)+g⋅(x−y) for all x in Rn. Therefore, g⋅y−f(y)≥g⋅x−f(x), so c⋅(y,s)>c⋅(x,t) , and c represents a separating hyperplane.

Usage

A strong separation oracle can be given as an input to the ellipsoid method for solving a linear program. Consider the linear program maximize  c⋅x  subject to  Ax≤b,x≥0. The ellipsoid method maintains an ellipsoid that initially contains the entire feasible domain Ax≤b. At each iteration t, it takes the center xt of the current ellipsoid, and sends it to the separation oracle:

After making a cut, we construct a new, smaller ellipsoid, that contains the remaining region. It can be shown that this process converges to an approximate solution, in time polynomial in the required accuracy.

Converting a weak oracle to a strong oracle

Given a weak separation oracle for a polyhedron, it is possible to construct a strong separation oracle by a careful method of rounding, or by diophantine approximations.: 159 

See also