Independent set (graph theory)

A set is independent if and only if it is a clique in the graph’s complement, so the two concepts are complementary. In fact, sufficiently large graphs with no large cliques have large independent sets, a theme that is explored in Ramsey theory.

A set is independent if and only if its complement is a vertex cover.[4] Therefore, the sum of the size of the largest independent set ${\displaystyle \alpha (G)}$ and the size of a minimum vertex cover ${\displaystyle \beta (G)}$ is equal to the number of vertices in the graph.

A vertex coloring of a graph ${\displaystyle G}$ corresponds to a partition of its vertex set into independent subsets. Hence the minimal number of colors needed in a vertex coloring, the chromatic number ${\displaystyle \chi (G)}$, is at least the quotient of the number of vertices in ${\displaystyle G}$ and the independent number ${\displaystyle \alpha (G)}$.

In a bipartite graph with no isolated vertices, the number of vertices in a maximum independent set equals the number of edges in a minimum edge covering; this is Kőnig's theorem.

An independent set that is not a proper subset of another independent set is called maximal. Such sets are dominating sets. Every graph contains at most 3^n/3 maximal independent sets,[5] but many graphs have far fewer. The number of maximal independent sets in n-vertex cycle graphs is given by the Perrin numbers, and the number of maximal independent sets in n-vertex path graphs is given by the Padovan sequence.[6] Therefore, both numbers are proportional to powers of 1.324718..., the plastic number.

The independent set problem and the clique problem are complementary: a clique in G is an independent set in the complement graph of G and vice versa. Therefore, many computational results may be applied equally well to either problem. For example, the results related to the clique problem have the following corollaries:

• The independent set decision problem is NP-complete, and hence it is not believed that there is an efficient algorithm for solving it.
• The maximum independent set problem is NP-hard and it is also hard to approximate.

Despite the close relationship between maximum cliques and maximum independent sets in arbitrary graphs, the independent set and clique problems may be very different when restricted to special classes of graphs. For instance, for sparse graphs (graphs in which the number of edges is at most a constant times the number of vertices in any subgraph), the maximum clique has bounded size and may be found exactly in linear time;[7] however, for the same classes of graphs, or even for the more restricted class of bounded degree graphs, finding the maximum independent set is MAXSNP-complete, implying that, for some constant c (depending on the degree) it is NP-hard to find an approximate solution that comes within a factor of c of the optimum.[8]

The maximum independent set problem is NP-hard. However, it can be solved more efficiently than the O(n² 2ⁿ) time that would be given by a naive brute force algorithm that examines every vertex subset and checks whether it is an independent set.

As of 2017 it can be solved in time O(1.1996ⁿ) using polynomial space.[9] When restricted to graphs with maximum degree 3, it can be solved in time O(1.0836ⁿ).[10]

For many classes of graphs, a maximum weight independent set may be found in polynomial time. Famous examples are claw-free graphs,[11] P₅-free graphs[12] and perfect graphs.[13] For chordal graphs, a maximum weight independent set can be found in linear time.[14]

Modular decomposition is a good tool for solving the maximum weight independent set problem; the linear time algorithm on cographs is the basic example for that. Another important tool are clique separators as described by Tarjan.[15]

Kőnig's theorem implies that in a bipartite graph the maximum independent set can be found in polynomial time using a bipartite matching algorithm.

In general, the maximum independent set problem cannot be approximated to a constant factor in polynomial time (unless P = NP). In fact, Max Independent Set in general is Poly-APX-complete, meaning it is as hard as any problem that can be approximated to a polynomial factor.[16] However, there are efficient approximation algorithms for restricted classes of graphs.

In planar graphs, the maximum independent set may be approximated to within any approximation ratio c < 1 in polynomial time; similar polynomial-time approximation schemes exist in any family of graphs closed under taking minors.[17]

In bounded degree graphs, effective approximation algorithms are known with approximation ratios that are constant for a fixed value of the maximum degree; for instance, a greedy algorithm that forms a maximal independent set by, at each step, choosing the minimum degree vertex in the graph and removing its neighbors, achieves an approximation ratio of (Δ+2)/3 on graphs with maximum degree Δ.[18] Approximation hardness bounds for such instances were proven in Berman & Karpinski (1999). Indeed, even Max Independent Set on 3-regular 3-edge-colorable graphs is APX-complete.[19]

An interval graph is a graph in which the nodes are 1-dimensional intervals (e.g. time intervals) and there is an edge between two intervals iff they intersect. An independent set in an interval graph is just a set of non-overlapping intervals. The problem of finding maximum independent sets in interval graphs has been studied, for example, in the context of job scheduling: given a set of jobs that has to be executed on a computer, find a maximum set of jobs that can be executed without interfering with each other. This problem can be solved exactly in polynomial time using earliest deadline first scheduling.

A geometric intersection graph is a graph in which the nodes are geometric shapes and there is an edge between two shapes iff they intersect. An independent set in a geometric intersection graph is just a set of disjoint (non-overlapping) shapes. The problem of finding maximum independent sets in geometric intersection graphs has been studied, for example, in the context of Automatic label placement: given a set of locations in a map, find a maximum set of disjoint rectangular labels near these locations.

Finding a maximum independent set in intersection graphs is still NP-complete, but it is easier to approximate than the general maximum independent set problem. A recent survey can be found in the introduction of Chan & Har-Peled (2012).

The problem of finding a maximal independent set can be solved in polynomial time by a trivial greedy algorithm.[20] All maximal independent sets can be found in time O(3^n/3) = O(1.4423ⁿ).