# Improved bounds for the greedy strategy in optimization problems with curvature

- 187 Downloads

## Abstract

Consider the problem of choosing a set of actions to optimize a real-valued polymatroid function subject to matroid constraints. The greedy strategy, an approximate solution, is known to satisfy some bounds in terms of the total curvature. The total curvature depends on function values on sets outside the constraint matroid. If the function is defined only on the matroid, the problem still makes sense, but the existing bounds involving the total curvature do not apply, which is puzzling. This motivates an alternative formulation of such bounds. The first question we address is whether it is possible to extend a polymatroid function defined on a matroid to one on the entire power set. This was recently shown to be negative in general. Here, we provide necessary and sufficient conditions for the existence of an *incremental* extension of a polymatroid function defined on the uniform matroid of rank *k* to one with rank \(k+1\), together with an algorithm for constructing the extension. Whenever a polymatroid function defined on a matroid can be extended to the entire power set, the bounds involving the total curvature of the extension apply. However, these bounds still depend on sets outside the constraint matroid. Motivated by this, we define a new notion of curvature called *partial curvature*, involving only sets in the matroid. We derive necessary and sufficient conditions for an extension to have a total curvature equal to the partial curvature. Moreover, we prove that the bounds in terms of the partial curvature are in general improved over the previous ones. We illustrate our results with two contrasting examples motivated by practical problems.

## Keywords

Curvature Greedy Matroid Polymatroid Submodular## Mathematics Subject Classification

90C27 90C59## Notes

### Funding

Funding was provided by Division of Computing and Communication Foundations (Grant No. CCF-1422658).

## References

- Boros E, Elbassioni K, Gurvich V, Khachiyan L (2003) An inequality for polymatroid functions and its applications. Discrete Appl Math 131(2):255–281MathSciNetCrossRefzbMATHGoogle Scholar
- Conforti M, Cornuéjols G (1984) Submodular set functions, matroids and the greedy algorithm: tight worst-case bounds and some generalizations of the Rado–Edmonds theorem. Discrete Appl Math 7:251–274MathSciNetCrossRefzbMATHGoogle Scholar
- Edmonds J (2003) Submodular functions, matroids, and certain polyhedra. Comb Optim 2570:11–26zbMATHGoogle Scholar
- Fisher ML, Nemhauser GL, Wolsey LA (1978) An analysis of approximations for maximizing submodular set functions—II. Math Program Stud 8:73–87MathSciNetCrossRefzbMATHGoogle Scholar
- Liu Y, Zhang Z, Chong EKP, Pezeshki A (2018) Performance bounds with curvature for batched greedy strategy. J Optim Theory Appl 177(2):535–562MathSciNetCrossRefzbMATHGoogle Scholar
- Nemhauser GL, Wolsey LA, Fisher ML (1978) An analysis of approximations for maximizing submodular set functions—I. Math Program Stud 14:265–294MathSciNetCrossRefzbMATHGoogle Scholar
- Singh AP, Guillory A, Bilmes J (2012) On bisubmodular maximization. In: Proceedings of 15th international conference on artificial intelligence and statistics, pp 1055–1063Google Scholar
- Streeter M, Golovin D (2008) An online algorithm for maximizing submodular functions. In: Proceedings of advances in neural information processing systems, pp 1577–1584Google Scholar
- Sviridenko M, Vondrák J, Ward J (2015) Optimal approximation for submodular and supermodular optimization with bounded curvature. In: Proceedings of 26th annual ACM-SIAM symposium. Discrete algorithms, pp 1134–1148Google Scholar
- Tutte WT (1965) Lecture on matroids. J Res Natl Bur Stand B Math Math Phys 69B:1–47MathSciNetCrossRefzbMATHGoogle Scholar
- Vondrák J (2010) Submodularity and curvature: the optimal algorithm. RIMS Kokyuroku Bessatsu B23:253–266MathSciNetzbMATHGoogle Scholar
- Zhang Z, Chong EKP, Pezeshki A, Moran W (2016) String submodular functions with curvature constraints. IEEE Trans Autom 61(3):601–616MathSciNetCrossRefzbMATHGoogle Scholar