# Locating tree-shaped facilities using the ordered median objective

## Abstract.

In this paper we consider the location of a tree-shaped facility *S* on a tree network, using the ordered median function of the weighted distances to represent the total transportation cost objective. This function unifies and generalizes the most common criteria used in location modeling, e.g., median and center. If there are *n* demand points at the nodes of the tree *T*=(*V*,*E*), this function is characterized by a sequence of reals, Λ=(*λ* _{1}, . . . ,*λ* _{ n }), satisfying *λ* _{1}≥*λ* _{2}≥...≥*λ* _{ n }≥0. For a given subtree *S* let *X*(*S*)={*x* _{1}, . . . ,*x* _{ n }} be the set of weighted distances of the *n* demand points from *S*. The value of the ordered median objective at *S* is obtained as follows: Sort the *n* elements in *X*(*S*) in nonincreasing order, and then compute the scalar product of the sorted list with the sequence Λ. Two models are discussed. In the tactical model, there is an explicit bound *L* on the length of the subtree, and the goal is to select a subtree of size *L*, which minimizes the above transportation cost function. In the strategic model the length of the subtree is variable, and the objective is to minimize the sum of the transportation cost and the length of the subtree. We consider both discrete and continuous versions of the tactical and the strategic models. We note that the discrete tactical problem is NP-hard, and we solve the continuous tactical problem in polynomial time using a Linear Programming (LP) approach. We also prove submodularity properties for the strategic problem. These properties allow us to solve the discrete strategic version in strongly polynomial time. Moreover the continuous version is also solved via LP. For the special case of the *k*-centrum objective we obtain improved algorithmic results using a Dynamic Programming (DP) algorithm and discretization and nestedness results.

## Keywords

Polynomial Time Transportation Cost Weighted Distance Demand Point Continuous Version## Preview

Unable to display preview. Download preview PDF.

## References

- 1.Alstrup, S., Lauridsen, P.W., Sommerlund, P., Thorup, M.: Finding cores of limited length. In: Algorithms and Data Structures, Lecture Notes in Computer Science n. 1271, Dehne, F., Rau-Chaplin, A. Sack, J-R., Tamassia R. (eds.), Springer, 1997, pp. 45–54Google Scholar
- 2.Becker, R.I., Perl, Y.: Finding the two-core of a tree. Discrete Appl. Math.
**11**, 103–113 (1985)MATHMathSciNetGoogle Scholar - 3.Becker, R.I., Lari, I., Scozzari, A.: Efficient algorithms for finding the (
*k*,*l*)-core of tree networks. Networks**40**, 208–215 (2003)CrossRefMATHGoogle Scholar - 4.Boffey, B., Mesa, J.A.: A review of extensive facility location in networks. European J. Operational Res.
**95**, 592–600 (1996)MATHGoogle Scholar - 5.Bramel, J., Simchi-Levi, D.: The Logic of Logistics: Theory, Algorithms and Applications for Logistics Management. Springer 1997, BerlinGoogle Scholar
- 6.Cho, G., Shaw, D.X.: A depth-first dynamic programming algorithm for the tree knapsack problem. INFORMS J. Comput.
**9**, 431–438 (1997)MathSciNetMATHGoogle Scholar - 7.Faigle, U., Kern, W.: Computational complexity of some maximum average weight problems with precedence constraints. Operations Res.
**42**, 688–693 (1994)MATHMathSciNetGoogle Scholar - 8.Fischetti, M., Hamacher, H.W., Jornsten, K., Maffioli, F.: Weighted
*k*-cardinality trees: complexity and polyhedral structure. Networks**24**, 11–21 (1994)MATHMathSciNetGoogle Scholar - 9.Francis, R.L., Lowe, T.J., Tamir, A.: Aggregation error bounds for a class of location models. Operations Res.
**48**, 294–307 (2000)CrossRefMATHMathSciNetGoogle Scholar - 10.Francis, R.L., Lowe, T.J., Tamir, A.: Worst-case incremental analysis for a class of
*p*-facility location problems. Networks**39**, 139–143 (2002)CrossRefMATHMathSciNetGoogle Scholar - 11.Groetschel, M., Lovasz, L., Schrijver, A.: Geometric Algorithms and Combinatorial Optimization. Springer-Verlag 1993, BerlinGoogle Scholar
- 12.Hakimi, S.L., Schmeichel, E.F., Labbe, M.: On locating path-or tree shaped facilities on networks. Networks
**23**, 543–555 (1993)MATHMathSciNetGoogle Scholar - 13.Hedetniemi, S.M., Cockaine, E.J., Hedetniemi, S.T.: Linear algorithms for finding the Jordan center and path center of a tree. Transportation Sci.
**15**, 98–114 (1981)CrossRefMathSciNetGoogle Scholar - 14.Johnson, D.S., Niemi, K.A.: On knapsack, partitions and a new dynamic technique for trees. Math. Operations Res.
**8**, 1–14 (1983)MATHGoogle Scholar - 15.Kalcsics, J., Nickel, S., Puerto, J.: Multi-facility ordered median problems: A further analysis. Networks
**41**, 1–12 (2003)MATHMathSciNetGoogle Scholar - 16.Kalcsics, J., Nickel, S., Puerto, J., Tamir, A.: Algorithmic results for ordered median problems defined on networks and the plane. Operations Res. Lett.
**30**, 149–158 (2002)MATHMathSciNetGoogle Scholar - 17.Kim, T.U., Lowe, T.J., Tamir, A., Ward, J.E.: On the location of a tree-shaped facility. Networks
**28**, 167–175 (1996)CrossRefMATHMathSciNetGoogle Scholar - 18.McCormick, S.T.: Submodular function minimization. To appear as a chapter in Handbook on Discrete Optimization, edited by K. Aardal, G. Nemhauser and R. WeismantelGoogle Scholar
- 19.Minieka, E.: Conditional centers and medians on a graph. Networks
**10**, 265–272 (1980)MathSciNetGoogle Scholar - 20.Minieka, E.: The optimal location of a path or tree in a tree network. Networks
**15**, 309–321 (1985)MATHMathSciNetGoogle Scholar - 21.Minieka, E., Patel, N.H.: On finding the core of a tree with a specified length. J. Algorithms
**4**, 345–352 (1983)MATHMathSciNetGoogle Scholar - 22.Morgan, C.A., Slater, J.P.: A linear algorithm for a core of a tree. J. Algorithms
**1**, 247–258 (1980)MATHMathSciNetGoogle Scholar - 23.Nickel, S., Puerto, J.: A unified approach to network location problems. Networks
**34**, 283–290 (1999)CrossRefMATHMathSciNetGoogle Scholar - 24.Ogryczak, W., Tamir, A.: Minimizing the sum of the
*k*largest functions in linear time. Inf. Processing Lett.**85**, 117–122 (2003)CrossRefMATHMathSciNetGoogle Scholar - 25.Peng, S., Lo, W.: Efficient algorithms for finding a core of a tree with specified length. J. Algorithms
**20**, 445–458 (1996)CrossRefMATHMathSciNetGoogle Scholar - 26.Peng, S., Stephens, A.B., Yesha, Y.: Algorithms for a core and
*k*-tree core of a tree. J. Algorithms**15**, 143–159 (1993)CrossRefMATHMathSciNetGoogle Scholar - 27.Queyranne, N.M.: Minimizing symmetric submodular functions. Math. Programming
**82**, 3–12 (1998)CrossRefMathSciNetGoogle Scholar - 28.Rodríguez-Chía, A.M., Nickel, S., Puerto, J., Fernández, F.R.: A flexible approach to location problems. Math. Meth. Oper. Res.
**51**, 69–89 (2000)CrossRefMATHGoogle Scholar - 29.Shioura, A., Shigeno, M.: The tree center problems and the relationship with the bottleneck knapsack problems. Networks
**29**, 107–110 (1997)CrossRefMATHMathSciNetGoogle Scholar - 30.Slater, P.J.: Locating central paths in a graph. Transportation Sci
**16**, 1–18 (1982)MathSciNetGoogle Scholar - 31.Tamir, A.: A unifying location model on tree graphs based on submodularity properties. Discrete Appl. Math.
**47**, 275–283 (1993)CrossRefMATHMathSciNetGoogle Scholar - 32.Tamir, A.: An
*O*(*pn*^{2}) algorithm for the*p*-median and related problems on tree graphs. Oper. Res. Lett.**19**, 59–64 (1996)CrossRefMATHMathSciNetGoogle Scholar - 33.Tamir, A.: Fully polynomial approximation schemes for locating a tree-shaped facility: a generalization of the knapsack problem. Discrete Appl. Math.
**87**, 229–243 (1998)CrossRefMATHMathSciNetGoogle Scholar - 34.Tamir, A.: The
*k*-centrum multi-facility location problem. Discrete Appl. Math.**109**, 292–307 (2000)MathSciNetGoogle Scholar - 35.Tamir, A.: Sorting weighted distances with applications to objective function evaluations in single facility location problems. Oper. Res. Lett.
**32**, 249–257 (2004)CrossRefMathSciNetMATHGoogle Scholar - 36.Tamir, A., Lowe, T.J.: The generalized
*p*-forest problem on a tree network. Networks**22**, 217–230 (1992)MathSciNetMATHGoogle Scholar - 37.Tamir, A., Puerto, J., Mesa, J.A., Rodriguez-Chia, A.M.: Conditional location of path and tree shaped facilities on trees. Technical Report, School of Mathematical Sciences, Tel-Aviv University, August 2001Google Scholar
- 38.Tamir, A., Puerto, J., Perez-Brito, D.: The centdian subtree on tree networks. Discrete Appl. Math.
**118**, 263–278 (2002)CrossRefMATHMathSciNetGoogle Scholar - 39.Tardos, E.: A strongly polynomial algorithm to solve combinatorial linear programs. Oper. Res.
**34**, 250–256 (1986)MATHMathSciNetGoogle Scholar - 40.Vaidya, P.M.: An algorithm for linear programming which requires
*O*((*m*+*n*)*n*^{2}+ (*m*+*n*)^{1.5}*nL*) arithmetic operations. Math. Programming**47**, 175–201 (1990)Google Scholar - 41.Wang, B-F: Efficient parallel algorithms for optimally locating a path and a tree of a specified length in a weighted tree network. J. Algorithms
**34**, 90–108 (2000)CrossRefMATHMathSciNetGoogle Scholar - 42.Wang, B-F: Finding a two-core of a tree in linear time. SIAM J. Discrete Math.
**15**, 193–210 (2002)CrossRefMATHMathSciNetGoogle Scholar