Multiple Facilities Location in the Plane Using the Gravity Model

Size: px
Start display at page:

Download "Multiple Facilities Location in the Plane Using the Gravity Model"

Transcription

1 Geographical Analysis ISSN Multiple Facilities Location in the Plane Using the Gravity Model Tammy Drezner, Zvi Drezner College of Business and Economics, California State University-Fullerton, Fullerton, CA Two problems are considered in this article. Both problems seek the location of p facilities. The first problem is the p median where the total distance traveled by customers is minimized. The second problem focuses on equalizing demand across facilities by minimizing the variance of total demand attracted to each facility. These models are unique in that the gravity rule is used for the allocation of demand among facilities rather than assuming that each customer selects the closest facility. In addition, we also consider a multiobjective approach, which combines the two objectives. We propose heuristic solution procedures for the problem in the plane. Extensive computational results are presented. Introduction The gravity model, first proposed by Reilly (1931) and later suggested for modeling market share in competitive situations by Huff (1964, 1966), is a spatial distribution model. When customers face a patronage choice among several facilities, it is commonly assumed (such as in location allocation models) that customers select the closest facility (Hotelling 1929). It can be argued that this distribution rule does not reflect reality unless there is central control that forces the selection of the closest facility. Customers may not have full information about distance to all facilities or may not behave in a rational manner. Furthermore, the proximity rule is discontinuous when the distances to the two closest facilities are comparable. In addition, if each demand point represents an area, the closest facility to some customers may be different than the closest facility to other customers in the same area. This phenomenon is termed as source C in aggregation error by Hillsman and Rhoda (1978). For a review of aggregation issues see Francis et al. (2004a, b, in press). According to the proximity rule, a facility is selected according to distances measured to the center of the area representing demand. Thus we can safely assume that not all customers at any location patronize the same facility. Correspondence: Zvi Drezner, College of Business and Economics, California State University-Fullerton, Fullerton, CA zdrezner@fullerton.edu Submitted: March 17, Revised version accepted: May 25, Geographical Analysis 38 (2006) r 2006 The Ohio State University 391

2 Geographical Analysis It is therefore more realistic to assume a spatial distribution of demand according to some probabilistic rule such as the gravity rule. This means that instead of assigning all customers at a demand point to the same facility (the closest one), customers are allocated to the competing facilities in inverse proportion to a nondecreasing function of the distance. The distance decay function can be a power of the distance (as suggested by Huff 1964, 1966), an exponential decay (Wilson 1976; Hodgson 1981), or other functional relationships (Bell, Ho, and Tang 1998). Drezner (1994a, b, 1995a) and Drezner and Drezner (1996) suggested alternative distributions of demand governed by a utility function, a random utility function and the gravity model. Spatial interaction and spatial choice models are described in detail with a comprehensive literature review in Fotheringham and O Kelly (1989) (see also Yamashita 1993). Such models are based on site attributes (independent of location), situation attributes (relative distances to alternative choices), and separation attributes (the distances between customers and facilities). Probabilistic approaches include the random utility maximization and the Logit model (Fotheringham and O Kelly 1989; Drezner and Drezner 1996; Drezner, Wesolowsky, and Drezner 1998). Fotheringham (1983, 1986) combined features of the random utility model and the gravity model. In the gravity model it is assumed that the probability of patronizing a facility (facility utility) is proportional to the attractiveness of the facility and inversely proportional to a function of the distance. Fotheringham calculated a similar expression using the random utility function. Both models, then, calculate the probability of patronage of a given facility by dividing the utility value calculated for that facility by the sum of the utility values for all competing facilities. The gravity rule was applied to hub selection (Drezner and Drezner 2001), to the p-median problem on a network (Drezner and Drezner in press), and to location in discrete space (Hodgson 1978; O Kelly and Storbeck 1984). Gravity-based models are suitable for situations where there is no central control and customers freely select a facility to patronize. When there is central control, such as in public schools, customers are assigned to the closest facility, thus satisfying the objective of minimizing the total distances traveled. In such cases, the proximity rule is the appropriate objective. When there is no central control, customers do not necessarily select the closest facility and a probabilistic rule such as the gravity-based model is the appropriate one. In public service-oriented applications such as post office or Department of Motor Vehicles branches, the gravity p-median model seeks to provide the best service in terms of driving distances while the gravity equity model seeks to evenly distribute demand, or equalize facility load, so that customers do not have to wait in long lines because one facility is busier than another. The p-median problem is well researched in the literature (see, e.g., on the network: Daskin 1995; Current, Daskin, and Schilling 2002, and in the plane: Love, Morris, and Wesolowsky 1988). The objective is to minimize the total travel distance (or the average distance) for all customers using the proximity rule. 392

3 Tammy Drezner and Zvi Drezner Multiple Facilities Location in the Plane In the equity model proposed in this article, equity is defined as equitable load distribution across facilities taking the facilities perspective rather than equidistance for customers. Minimizing the variance of loads is achieved by calculating the average load, which is independent of the locations of the facilities (it is the total demand divided by the number of facilities), and minimizing the sum of squares of deviations from this average. Attempting to find a distribution of loads with small deviations from the average was suggested in Church and Murray (1993). The equity model is reminiscent of the voting districting problem (Hess et al. 1965; Garfinkel and Nemhauser 1970; Meholtra, Johnson, and Nemhauser 1998) and to the sales territories delineation problem (Kalcsics, Melo, and Nickel 2002). In the voting districting problem districts need to be delineated so as to equalize the number of voters across districts. In the sales territories problem the territories need to be of comparable buying power or sales potential. In both the voting districting and the sales territories problems the districts need to be connected and form simple shapes. The gravity model is inappropriate for modeling these two problems because there is a central control determining where each voter votes or which customers are assigned to each territory. Models involving minimization of the variance of distances traveled by customers are addressed on the network in Maimon (1986) and in the plane in Carrizosa (1997, 1999) and Drezner, Drezner, and Salhi (2006). The equity model is closely related to the capacitated p-median (van Roy 1986) or p-center problems (Mirchandani and Francis 1990; Drezner 1995b). The equity model is equivalent to finding the lowest capacity, which yields a feasible solution to the capacitated problem. For a discussion of equity models see Marsh and Schilling (1994), Eiselt and Laporte (1995), Berman, Drezner, and Wesolowsky (2003), and Drezner (2004). In the current article, we consider location of multiple facilities in the plane using Euclidean distances. Three problems are analyzed: (a) the p-median (location allocation) problem in the plane using the gravity model, (b) the equity model in the plane using the gravity model (analyzed in a network environment in Berman, Drezner, and Wesolowsky 2003) where the objective is to distribute demand as uniformly as possible among the facilities, and (c) the multiobjective model which incorporates both objectives. The multiobjective approach finds a solution close to the p-median solution (which has the minimal possible cost), while increasing the cost by a relatively small amount to achieve a more equitable solution. Another approach to satisfying both objectives in consolidation of public schools was suggested by Diamond and Wright (1987) who suggested a multiobjective model and Church and Murray (1993) who pointed out that the Diamond and Wright (1987) model tends to recommend closing small schools which is inequitable and propose an improved algorithm to eliminate this bias. In one of their models they attempt to minimize the deviations of the individual utilization values from the average utilization, which is the same concept employed by the variance minimization model. 393

4 Geographical Analysis The solution to the equity model presented in this article tends to be dispersed in the region, which is similar in many ways to the p-dispersion or obnoxious facilities models. In the p-dispersion model (Kuby 1987) we wish to locate p facilities in a bounded region such that the minimum distance between any two facilities is maximized. In obnoxious facility models (Church and Garfinkel 1978; Hansen, Peeters, and Thisse 1981) we wish to locate one or more facilities that are as far as possible from a given set of demand points. The article is organized as follows: in the next section, we present the basic formulas for the three models. In the following section, the models are formulated and in section that follows, solution procedures are presented. In the penultimate section computational experiments are reported and we conclude in the last section. Notation and basic formulas For flexibility in the modeling we define for each facility an attractiveness level, and a capacity (or factor to determine equity). For most applications both the attractiveness level and the capacity factor are assumed to be equal to one. In the classical p-median model, for example, all facilities are equally attractive and have equal capacity. Equal attractiveness is implicitly assumed in the proximity rule. Let: n be the number of demand points p be the number of facilities (a i, b i ) be the location of demand point i for i 5 1,..., n (x j, y j ) be the unknown location of facility j for j 5 1,...,p X be the vector of locations (x i, y i ), for j 5 1,..., p d ij be the distance between demand point i and facility j for i 5 1,..., n; j 5 1,..., p w i be the demand generated at demand point i for i 5 1,..., n A j be the attractiveness of facility j for j 5 1,..., p (normally equal to 1) a j be the factor (inverse of capacity) of facility j for j 5 1,..., p (normally equal to 1) l be the power to which the distance is raised By the gravity rule: The proportion of the demand generated at demand point i that patronizes facility k is A k d r¼1 A rd ir ð1þ 394

5 Tammy Drezner and Zvi Drezner Multiple Facilities Location in the Plane The average distance traveled by a customer residing at demand point i is A k d P k¼1 p d r¼1 A r d ir P ¼ Xp p A k d P k¼1 p r¼1 A r d ir k¼1 The total demand served by facility k is w i A k d r¼1 A rd ir A k d r¼1 A rdir d ¼ k¼1 A kd 1 k¼1 A kd ð2þ ð3þ The models Three models are analyzed: The p-median model, the minimum variance model, and a multiobjective model, which incorporates these two models. The location allocation (p-median) model In the location allocation (p-median) model customers total travel distance is minimized. By equation (2) the total distance T(X) traveled is P T ðxþ ¼ Xn p k¼1 w A kd 1 i k¼1 A ð4þ kd The gradient of T(X) is calculated in Appendix A. In Drezner and Drezner (in press) several properties of the gravity p-median model on the network were proven. The same properties are also valid for the gravity p-median model in the plane. The proofs are based on the property that the weighted mean of distances is greater than or equal to the minimum distance with equality holding for p 5 1. The properties are: Property 1. The value of the objective function of the standard p-median problem is not higher than the value of the objective function for the gravity p-median problem. Property 2. The gravity 1-median problem is identical to the standard 1-median problem. Property 1 demonstrates that central control leads to a better value of the objective function. Property 2 implies that, when a location of one facility is sought, the two models result in the same location solution. Minimum variance model The total demand patronizing all facilities is constant: P n w i. If all loads are proportional to the factor a i, the ideal demand allocated to facility k is a k = r¼1 a r P n w i (or 1=p P n w i for equal factors). Minimizing the sum of squares of 395

6 Geographical Analysis deviations of the actual demand from the desired demand is (and defining F(X) and F k (X)): FðXÞ ¼ Xp k¼1 w i A k d r¼1 A rdir a k r¼1 a r w i! 2 ¼ Xp k¼1 F k ðxþ ð5þ Note that the variance of the demand loads at the facilities, for equal factors, is F(X)/p or F(X)/(p 1) for sample variance. The gradient of F(X) is calculated in Appendix A. For the minimum variance problem there are two types of optimum with a zero variance: Type 1. If all facilities are located at the same point, demand is evenly distributed among all facilities and thus the variance of the allocated demand is zero. Type 2. If all facilities are located at infinity, demand is also evenly distributed among the facilities with a zero variance. We experimented extensively with the gradient search to solve the minimum variance problem and repeatedly obtained a zero variance solution. Examining the solutions obtained we found that in all cases, the solution was a Type 2 one, that is, the facilities were located at a great distance from the set of demand points. The gradient appears to push the facilities away from the cluster of demand points rather than attract them to a common location. Such a solution is a poor location solution in terms of the quality of service provided by the facilities. The average distance customers have to travel to the facilities is unnecessarily excessive. Similar results for the Lorenz equity measure were obtained by Drezner (2004) for the location of casualty collection points. Note that the Type 1 optimal solution where all facilities are located at one point is not a good location solution either. The best Type 1 solution is to locate all facilities at the 1-median solution. If the 1-median location solution is used, the advantage of having multiple facilities is lost, and one might as well utilize only one facility. It follows that, in order to obtain equity, all customers receive poor service. A multiobjective approach As the minimum variance problem is not an interesting one by itself, we investigated the multiobjective of the p median combined with the minimum variance. We define the objective function of H(X) 5 T(X) 5 yf(x) for some y40. The gradient of H(X) is simply HHðXÞ ¼HT ðxþþyhf ðxþ 396

7 Tammy Drezner and Zvi Drezner Multiple Facilities Location in the Plane Objective Function diagonal axis Distance from Center Figure 1. The value of the objective function for the square example. Solution procedure We solve the three models by using the gradient search method, a technique commonly used for the optimization of an objective function with no constraints in a continuous space (see Appendix A for details). Running the gradient search on the p-median (or the multiobjective) model we observed cases where the number of iterations was vary large, occasionally exceeding 100,000, compared with about 100 iterations for most problems. Examining these special cases we found that at least one facility was located at a great distance from the cluster of demand points and its progress according to the gradient was very slow. As all facilities were generated in the convex hull of the demand points, we were puzzled as to the reason for this phenomenon. To better understand the reasons for this phenomenon, we analyzed a case of locating one facility to serve a square of side 2 centered at the origin with uniform demand and a second facility located at the origin. The objective function is the p median. Details of the analysis are given in Appendix B. An examination of Fig. 1 reveals that the value of the p-median objective function decreases when the second facility is located farther than approximately 2.5 units from the center of the square. It is plausible that under different configurations, the decrease in the value of the objective function starts inside the square. This analysis provides an explanation to the rare occurrences in which one facility is moved away from the square to infinity causing the excessive number of iterations, which in turn, leads to a lengthy computational effort and inferior solutions. Note that this phenomenon is not an issue in the standard p-median problem where the value of the objective function cannot decrease when a facility is located farther away from the convex hull of demand points. In fact, once a facility is farther 397

8 Geographical Analysis from the boundary of the convex hull than the diameter of the convex hull, no demand point is using its services and the objective function remains constant beyond that point. As a result of this analysis, we implemented the following modification to the gradient search: in every iteration it is checked whether a facility is located outside the convex hull. If this is encountered: (i) (ii) (iii) the present solution is discarded, all the facilities are randomly generated inside the convex hull (yielding a completely new starting solution), and the iterations continue. With this modification, we did not experience any instances of excessively long run times. Generating dispersed starting solutions As there is no real advantage in locating two facilities next to one another, we expect the solution to both the p-median and the multiobjective models to be dispersed. Therefore, we first experimented with generating dispersed starting solutions following the approach used in Drezner, Drezner, and Salhi (2002). A parameter K is used. The location for the first facility is randomly generated. We repeat the following for the location of the rest of the facilities: K candidate locations are randomly generated; the candidate location, which is farthest from the already selected locations is chosen as the starting location for the facility. Note that K 5 1 is the standard way of generating starting solutions. In Table 1 we report the results for K 5 1,..., 10, with each instance of K solved 10 times from 10 randomly generated starting solutions. The best solution found for each problem for all 10 K s is the best found (BF) solution. This BF solution may be improved in subsequent experiments to the best known (BK) solution. Table 1 depicts: (i) (ii) (iii) (iv) the number of problems for which the BF solution was found at least once, the average percent of the BF solution for each K over the BF for the 44 problems, the average percent of the average solution for each K over the BF for all 44 problems, and the total time in hours it took to solve all 44 problems 10 times each. We selected for the experiments K 5 4 which seems to perform best (see Table 1). Computational experiments Fortran programs were coded, compiled by Microsoft Fortran PowerStation 4.0, and were run on a 2.8 MHz Pentium IV computer with 256 Mb RAM. Problems were randomly generated in a unit square, with w i 5 1, i 5 1,..., n, A j 5 1, j 5 1,...,p, and a j 5 1, j 5 1,...,p. Problems were generated for n 5 10, 50, 100, 1000, 398

9 Tammy Drezner and Zvi Drezner Multiple Facilities Location in the Plane Table 1 Results for Various Values of K Property K 5 1 K 5 2 K 5 3 K 5 4 K 5 5 K 5 6 K 5 7 K 5 8 K 5 9 K 5 10 The p-median problem # BF (of 44) % over BF % of average over BF Total runtime (h) The multiobjective problem # BF (of 44) % over BF % of average over BF Total runtime (h)

10 Geographical Analysis Table 2 Best-Known Results N p p-median Multiobject N p p-median Multiobject , , , , , , , , , and 10,000 demand points. The number of facilities to be located is p 5 2, 3, 4, 5, 10, 20, 50, 100 with the stipulation that 5p n, yielding 44 different problems. We did not experiment with the location of one facility because the 1-median in the plane is a convex problem. Note that one set of demand points was generated for each n and was used for all p s with the same n. If during the iterations a facility is located outside the square, the solution is abandoned and a new random starting solution is generated. For the multiobjective model we normalized the variance objective by dividing F(X) by the mean weight n/p, that is, using y 5 p/n. This normalization borrows from the formula for the w 2 test. We ran each problem 100 times. In Table 2 we report the BK solution for the gravity p-median and the multiobjective models for all 44 problems incorporating the best results found in Table 1. Table 3 depicts the results for each of the 44 problems, each run 100 times using K 5 4. For each of the two objectives, the p-median and the multiobjective, we report: (a) the number of times the BK solution was found, (b) the percentage of the BF solution over the BK solution, (c) the average percentage over the BK 400

11 Tammy Drezner and Zvi Drezner Multiple Facilities Location in the Plane Table 3 Results for both objectives on the test problems N p The p-median The multiobjective # best % best % average Time (s) # best % best % average Time (s) , , , , , , , , , , , , , , , ,

12 Geographical Analysis solution, and (d) the total run time in seconds for running all 100 replications for each problem. Results For the gravity p-median problem, the BK solution was found for 41 of 44 problems. The average of the BF solution over the BK solution is %. The average solution was 0.172% over the BK solution. Total time for all 44 problems, each solved 100 times, is h. For the multiobjective problem, the BK solution was found for 43 of 44 problems. The average of the BF solution is practically the same as the BK solution. The average solution was 0.088% over the BK solution. Total time for all 44 problems, each solved 100 times, is h. In conclusion, applying K 5 4 for generating dispersed starting solutions performed very well. Run times for gravity multiobjective problems are about four to five times slower than those required for the solution of gravity p-median problems. The average of the solutions did not exceed 2% above the BK solution for any of the problems. The gravity multiobjective model provided average results, which are about twice as close to the BK solution than the gravity p-median results. Conclusions In this article we considered the p-median, minimum variance of loads, and the combined multiobjective location models. The objective of the p-median model is to minimize the total cost for customers. The objective of the minimum load variance is to provide an equitable solution for the facilities. Both of these objectives are important and therefore we proposed the multiobjective model. The goal of the multiobjective approach is to consider both objectives and find a solution where the total cost for customers is a bit higher than the minimum possible cost (obtained at the p-median solution) while improving the equity by reducing the variance from its value at the p-median solution. Allocation of demand to facilities is estimated by the gravity rule. In many applications customers do not necessarily use the services of the closest facility to them. The gravity rule divides the customers located at each demand point among all facilities. The gravity rule allocates more customers to closer facilities and only a few of them to far away facilities. A gradient search was constructed for the solution of these problems with very good results. We also incorporated the construction of dispersed starting solutions and analyzed the effect of the construction of the starting solution on the final result of the algorithm. As future research we suggest to solve the multiobjective model proposed in this model in a network environment. 402

13 Tammy Drezner and Zvi Drezner Multiple Facilities Location in the Plane Appendix A Gradient search The following is the standard gradient search. To simplify the notation we used H(X) in the description of the algorithm that stands for either T(X) or F(X). 1. Generate p random locations for the facilities: (x j, y j ) for j 5 1,..., p. 2. Calculate the gradient (dh(x)/dx j, dh(x)/dy j for j 5 1,..., p). 3. If the maximum absolute value of the elements of the gradient is less than a prespecified e, stop with the present (x j, y j ) as the solution. 4. Perform a one-dimensional search on t 0 by the Golden Section search (Zangwill 1969). The objective function to be minimized is H(x j tdh(x)/dx j, y j tdh (X)/dy j for j 5 1,...,p). 5. Update (x j, y j ) by the best t and go to step 2. Calculating the gradient of T(X) We calculate the derivative dt(x)/dx j for 1 j p. dtðxþ dx j ¼ Xn ¼ A j w i k¼1 A kd 2 ð1 lþa j dij ( X p X ) p A k d þ la j dij 1 A k d 1 xj a i d k¼1 k¼1 ij ( ) w i dij 2 ðx j a i Þ X p k¼1 A kd 2 ð1 lþd ij A k d þ l Xp A k d 1 k¼1 k¼1 ð6þ and a similar expression holds for dt(x)/dy j. Calculating the gradient of F(X) We first find the derivative F k (X)byx j for some j. This yields different expressions for j 5 k and j6¼k. For j 5 k ¼ 2 k w i ¼ 2lw i A k d 1 A k d r¼1 A rdir w i a k r¼1 a r A k d r¼1 A rd ir! A k d 1 r¼1 w i w A rdir þ la 2 k d 2l 1 i r¼1 A rdir 2 a k r¼1 a r! r¼16¼k w A rdir i r¼1 A rdir 2 403

14 Geographical Analysis For k ij ¼ 2 Xn w i A k d r¼1 A k ðxþ ¼ 2 lw ia j d 1 X ij ij r¼1 A rd 2 8 >< >: r¼16¼k ir A r d ir A k d a k r¼1 a r jj ¼ k jj 6¼ k w i A k d r¼1 A rd ir! A k d w i w d ij 1 i r¼1 A rdir 2! a k r¼1 a r w i ð7þ According to the chain rule: or df k ðxþ dx j ¼ k ðxþ x j a ij d ij dfðxþ dx j ¼ Xp k ðxþ x j a ij d ij ð8þ and similarly dfðxþ dy j ¼ Xp k ðxþ y j b ij d ij Appendix B Location of two facilities in a uniform square demand To analyze this special case, let one facility be located at the center of the square, and the other facility be located at a point (a, b). We considered two locations for (a, b): (d, 0)or(d, d), and used l 5 2. The value of the p-median objective function is by (4): T ða; bþ ¼ Z 1 Z p 1 þ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffiffiffiffiffiffiffiffiffiffi 1 x 2 þy 2 ðx aþ 2 þðy bþ 2 dxdy ð9þ 1 1 þ x 2 þy 2 ðx aþ 2 þðy bþ 2 The value of the objective function in (9) was calculated using the Simpson rule for each dimension. The value of the objective function as a function of the distance from the origin for (a, b) 5 (d, 0) (on the axis) and (a, b) 5 (d, d) (on the diagonal) is depicted in Fig

15 Tammy Drezner and Zvi Drezner Multiple Facilities Location in the Plane References Bell, D. R., T.-H. Ho, and C. S. Tang. (1998). Determining Where to Shop: Fixed and Variable Costs of Shopping. Journal of Marketing Research 35, Berman, O., Z. Drezner, and G. O. Wesolowsky. (2003). Optimal Location With Equity, under review. Carrizosa, E. (1997). Locating a Desirable Facility With an Equity Criterion. Studies in Locational Analysis 11, Carrizosa, E. (1999). Minimizing the Variance of Euclidean Distances. Studies in Locational Analysis 12, Church, R. L., and R. S. Garfinkel. (1978). Locating an Obnoxious Facility on a Network. Transportation Science 12, Church, R. L., and A. T. Murray. (1993). Modeling School Utilization and Consolidation. Journal of Urban Planning and Development 119, Current, J., M. Daskin, and D. Schilling. (2002). Discrete Network Location Models. In Location Analysis: Applications and Theory, , edited by Z. Drezner and H. W. Hamacher. Berlin, Germany: Springer. Daskin, M. S. (1995). Network and Discrete Location: Models, Algorithms, and Applications. New York: Wiley. Diamond, J. T., and J. R. Wright. (1987). Multiobjective Analysis of Public School Consolidation. Journal of Urban Planning and Development 113, Drezner, T. (1994a). Locating a Single New Facility Among Existing Unequally Attractive Facilities. Journal of Regional Science 34, Drezner, T. (1994b). Optimal Continuous Location of a Retail Facility, Facility Attractiveness, and Market Share: An Interactive Model. Journal of Retailing 70, Drezner, T. (1995a). Competitive Facility Location in the Plane. In Facility Location: A Survey of Applications and Methods, , edited by Z. Drezner. New York: Springer. Drezner, T. (2004). Location of Casualty Collection Points. Environment and Planning C 22, Drezner, Z., ed. (1995b). Facility Location: A Survey of Applications and Methods. New York: Springer. Drezner, T., and Z. Drezner. (1996). Competitive Facilities: Market Share and Location with Random Utility. Journal of Regional Science 36, Drezner, T., and Z. Drezner. (2001). A Note on Applying the Gravity Rule to the Airline Hub Problem. Journal of Regional Science 41, Drezner, T., and Z. Drezner. (in press). The Gravity p-median Model. European Journal of Operational Research. Drezner, T., Z. Drezner, and S. Salhi. (2002). Solving the Multiple Competitive Facilities Location Problem. European Journal of Operational Research 142, Drezner, T., Z. Drezner, and S. Salhi. (2006). A Multi-Objective Heuristic Approach for the Casualty Collection Points Location Problem, The Journal of the Operational Research Society 58, Drezner, Z., G. O. Wesolowsky, and T. Drezner. (1998). On the Logit Approach to Competitive Facility Location. Journal of Regional Science 38, Eiselt, H. A., and G. Laporte. (1995). Objectives in Locations Problems. In Facility Location: A Survey of Applications and Methods, , edited by Z. Drezner. New York: Springer. Fotheringham, A. S. (1983). A New Set of Spatial Interaction Models: The Theory of Competing Destinations. Environment and Planning A 15, Fotheringham, A. S. (1986). Modeling Hierarchical Destination Choice. Environment and Planning A 18,

16 Geographical Analysis Fotheringham, A. S., and M. E. O Kelly. (1989). Spatial Interaction Models: Formulations and Applications. Dordrecht, Germany: Kluwer Academic Publishers. Francis, R. L., T. J. Lowe, B. Rayco, and A. Tamir. (in press). Aggregation Error for Location Models: Survey and Analysis, Annals of Operational Research. Francis, R. L., T. J. Lowe, A. Tamir, and H. Emir-Farinas. (2004a). Aggregation Decomposition and Aggregation Guidelines for a Class of Minimax and Covering Location Models. Geographical Analysis 36, Francis, R. L., T. J. Lowe, A. Tamir, and H. Emir-Farinas. (2004b). A Framework for Demand Point and Solution Space Aggregation Analysis for Location Models. European Journal of Operational Research 159, Garfinkel, R. S., and G. L. Nemhauser. (1970). Optimal Political Districting by Implicit Enumeration Techniques. Management Science 16(8), B Hansen, P., D. Peeters, and J.-F. Thisse. (1981). On the Location of an Obnoxious Facility. Sistemi Urbani 3, Hess, S., J. Weaver, H. Siegfeldt, J. Whelan, and P. Zitlau. (1965). Non Partisan Political Redistricting by Computer. Operations Research 13, Hillsman, E. L., and R. Rhoda. (1978). Errors in Measuring Distances from Populations to Service Centers. Annals of Regional Science 12, Hodgson, M. J. (1978). Towards More Realistic Allocation in Location Allocation Models: An Interaction Approach. Environment and Planning A 10, Hodgson, J. M. (1981). A Location Allocation Model Maximizing Consumers Welfare. Regional Studies 15, Hotelling, H. (1929). Stability in Competition. Economic Journal 39, Huff, D. L. (1964). Defining and Estimating a Trade Area. Journal of Marketing 28, Huff, D. L. (1966). A Programmed Solution for Approximating an Optimum Retail Location. Land Economics 42, Kalcsics, J., T. Melo, and S. Nickel. (2002, June). Planning Sales Territories A Facility Location Approach, ISOLDE IX Conference, Fredericton, New Brunswick, Canada. Kuby, M. (1987). Programming Models for Facility Dispersion: The p-dispersion and Maxisum Dispersion Problems. Geographical Analysis 19, Love, R. F., J. G. Morris, and G. O. Wesolowsky. (1988). Facilities Location: Models and Methods. Amsterdam: North Holland. Maimon, O. (1986). The Variance Equity Measure in Locational Decision Theory. Annals of Operations Research 6, Marsh, M., and D. Schilling. (1994). Equity Measurement in Facility Location Analysis: A Review and Framework. European Journal of Operational Research 74, Meholtra, A., E. L. Johnson, and G. L. Nemhauser. (1998). An Optimization Based Heuristic for Political Districting. Management Science 44, Mirchandani, P. B., and R. L. Francis, eds. (1990). Discrete Location Theory. New York: Wiley-Interscience. O Kelly, M. E., and J. E. Storbeck. (1984). Hierarchical Location Models with Probabilistic Allocation. Regional Studies 18, Reilly, W. J. (1931). The Law of Retail Gravitation. New York: Knickerbocker Press. van Roy, T. J. (1986). A Cross Decomposition Algorithm for Capacitated facility Location. Operations Research 34, Wilson, A. G. (1976). Retailers Profits and Consumers Welfare in a Spatial Interaction Shopping Model. In Theory and Practice in Regional Science, 42 59, edited by I. Masser. London: Pion. Yamashita, J. (1993). Effects of Spatial Interaction on Spatial Structure: A Case of Day Center Location in Malmo. Geographical Review of Japan 66, Zangwill, W. I. (1969). Nonlinear Programming: A Unified Approach. Englewood Cliffs, NJ: Prentice-Hall. 406

Institut für Numerische und Angewandte Mathematik

Institut für Numerische und Angewandte Mathematik Institut für Numerische und Angewandte Mathematik When closest is not always the best: The distributed p-median problem Brimberg, J.; Schöbel, A. Nr. 1 Preprint-Serie des Instituts für Numerische und Angewandte

More information

A national-scale application of the Huff gravity model for the estimation of town centre retail catchment area

A national-scale application of the Huff gravity model for the estimation of town centre retail catchment area A national-scale application of the Huff gravity model for the estimation of town centre retail catchment area Michail Pavlis, Les Dolega, Alex Singleton Department of Geography and Planning, School of

More information

The Trade Area Analysis Model

The Trade Area Analysis Model The Trade Area Analysis Model Trade area analysis models encompass a variety of techniques designed to generate trade areas around stores or other services based on the probability of an individual patronizing

More information

Models and Algorithms for Facility Location Problems with Equity Considerations

Models and Algorithms for Facility Location Problems with Equity Considerations University of Naples Federico II PhD School in Industrial Engineering Department of Industrial Engineering PhD Program in Science and Technology Management XXV Cycle PhD Thesis Models and Algorithms for

More information

Locating a competitive facility in the plane with a robustness criterion

Locating a competitive facility in the plane with a robustness criterion Journal to be determined manuscript No. (will be inserted by the editor) Locating a competitive facility in the plane with a robustness criterion R. Blanquero 1, E. Carrizosa 1, E.M.T. Hendrix 2 1 Facultad

More information

Oded BermanTammy DreznerZvi DreznerDmitry Krass

Oded BermanTammy DreznerZvi DreznerDmitry Krass This article was downloaded by: [148.251.232.83] On: 18 January 2019, At: 06:08 Publisher: Institute for Operations Research and the Management Sciences (INFORMS) INFORMS is located in Maryland, USA INFORMS

More information

Facility Location Problems with Random Demands in a Competitive Environment

Facility Location Problems with Random Demands in a Competitive Environment Facility Location Problems with Random Demands in a Competitive Environment Takeshi Uno, Hideki Katagiri, and Kosuke Kato Abstract This paper proposes a new location problem of competitive facilities,

More information

HUB NETWORK DESIGN MODEL IN A COMPETITIVE ENVIRONMENT WITH FLOW THRESHOLD

HUB NETWORK DESIGN MODEL IN A COMPETITIVE ENVIRONMENT WITH FLOW THRESHOLD Journal of the Operations Research Society of Japan 2005, Vol. 48, No. 2, 158-171 2005 The Operations Research Society of Japan HUB NETWORK DESIGN MODEL IN A COMPETITIVE ENVIRONMENT WITH FLOW THRESHOLD

More information

Equality Measures Properties for Location Problems

Equality Measures Properties for Location Problems Equality Measures Properties for Location Problems Maria Barbati Carmela Piccolo October 16, 2015 Abstract The objectives underlying location decisions can be various. Among them, equity objectives have

More information

A Tabu Search Approach based on Strategic Vibration for Competitive Facility Location Problems with Random Demands

A Tabu Search Approach based on Strategic Vibration for Competitive Facility Location Problems with Random Demands A Tabu Search Approach based on Strategic Vibration for Competitive Facility Location Problems with Random Demands Takeshi Uno, Hideki Katagiri, and Kosuke Kato Abstract This paper proposes a new location

More information

Minmax regret 1-center problem on a network with a discrete set of scenarios

Minmax regret 1-center problem on a network with a discrete set of scenarios Minmax regret 1-center problem on a network with a discrete set of scenarios Mohamed Ali ALOULOU, Rim KALAÏ, Daniel VANDERPOOTEN Abstract We consider the minmax regret 1-center problem on a general network

More information

The Journal of Database Marketing, Vol. 6, No. 3, 1999, pp Retail Trade Area Analysis: Concepts and New Approaches

The Journal of Database Marketing, Vol. 6, No. 3, 1999, pp Retail Trade Area Analysis: Concepts and New Approaches Retail Trade Area Analysis: Concepts and New Approaches By Donald B. Segal Spatial Insights, Inc. 4938 Hampden Lane, PMB 338 Bethesda, MD 20814 Abstract: The process of estimating or measuring store trade

More information

A Hub Location Problem with Fully Interconnected Backbone and Access Networks

A Hub Location Problem with Fully Interconnected Backbone and Access Networks A Hub Location Problem with Fully Interconnected Backbone and Access Networks Tommy Thomadsen Informatics and Mathematical Modelling Technical University of Denmark 2800 Kgs. Lyngby Denmark tt@imm.dtu.dk

More information

Operations Research Letters. A generalized Weiszfeld method for the multi-facility location problem

Operations Research Letters. A generalized Weiszfeld method for the multi-facility location problem Operations Research Letters 38 ) 7 4 Contents lists available at ScienceDirect Operations Research Letters journal homepage: www.elsevier.com/locate/orl A generalized Weiszfeld method for the multi-facility

More information

Shortening Picking Distance by using Rank-Order Clustering and Genetic Algorithm for Distribution Centers

Shortening Picking Distance by using Rank-Order Clustering and Genetic Algorithm for Distribution Centers Shortening Picking Distance by using Rank-Order Clustering and Genetic Algorithm for Distribution Centers Rong-Chang Chen, Yi-Ru Liao, Ting-Yao Lin, Chia-Hsin Chuang, Department of Distribution Management,

More information

Chapter 12. Key Issue Two: Why are consumer services distributed in a regular pattern?

Chapter 12. Key Issue Two: Why are consumer services distributed in a regular pattern? Chapter 12 Key Issue Two: Why are consumer services distributed in a regular pattern? Distribution of Consumer Services Central place theory Market area of a service Size of market area Market area analysis

More information

Facility Location Problem for Emergency and On-Demand Transportation Systems. Burak Boyacı Nikolas Geroliminis

Facility Location Problem for Emergency and On-Demand Transportation Systems. Burak Boyacı Nikolas Geroliminis On-Demand Transportation Systems Burak Boyacı Nikolas Geroliminis STRC 22 January 22 On-Demand Transportation Systems January 22 STRC 22 Facility Location Problem for Emergency and On-Demand Transportation

More information

Barnali Gupta Miami University, Ohio, U.S.A. Abstract

Barnali Gupta Miami University, Ohio, U.S.A. Abstract Spatial Cournot competition in a circular city with transport cost differentials Barnali Gupta Miami University, Ohio, U.S.A. Abstract For an even number of firms with identical transport cost, spatial

More information

The Newton Bracketing method for the minimization of convex functions subject to affine constraints

The Newton Bracketing method for the minimization of convex functions subject to affine constraints Discrete Applied Mathematics 156 (2008) 1977 1987 www.elsevier.com/locate/dam The Newton Bracketing method for the minimization of convex functions subject to affine constraints Adi Ben-Israel a, Yuri

More information

Włodzimierz Ogryczak. Warsaw University of Technology, ICCE ON ROBUST SOLUTIONS TO MULTI-OBJECTIVE LINEAR PROGRAMS. Introduction. Abstract.

Włodzimierz Ogryczak. Warsaw University of Technology, ICCE ON ROBUST SOLUTIONS TO MULTI-OBJECTIVE LINEAR PROGRAMS. Introduction. Abstract. Włodzimierz Ogryczak Warsaw University of Technology, ICCE ON ROBUST SOLUTIONS TO MULTI-OBJECTIVE LINEAR PROGRAMS Abstract In multiple criteria linear programming (MOLP) any efficient solution can be found

More information

No EFFICIENT LINE SEARCHING FOR CONVEX FUNCTIONS. By E. den Boef, D. den Hertog. May 2004 ISSN

No EFFICIENT LINE SEARCHING FOR CONVEX FUNCTIONS. By E. den Boef, D. den Hertog. May 2004 ISSN No. 4 5 EFFICIENT LINE SEARCHING FOR CONVEX FUNCTIONS y E. den oef, D. den Hertog May 4 ISSN 94-785 Efficient Line Searching for Convex Functions Edgar den oef Dick den Hertog 3 Philips Research Laboratories,

More information

THE NEWTON BRACKETING METHOD FOR THE MINIMIZATION OF CONVEX FUNCTIONS SUBJECT TO AFFINE CONSTRAINTS

THE NEWTON BRACKETING METHOD FOR THE MINIMIZATION OF CONVEX FUNCTIONS SUBJECT TO AFFINE CONSTRAINTS THE NEWTON BRACKETING METHOD FOR THE MINIMIZATION OF CONVEX FUNCTIONS SUBJECT TO AFFINE CONSTRAINTS ADI BEN-ISRAEL AND YURI LEVIN Abstract. The Newton Bracketing method [9] for the minimization of convex

More information

Ordered median problem with demand distribution weights

Ordered median problem with demand distribution weights Optim Lett (2016) 10:1071 1086 DOI 10.1007/s11590-015-0990-4 ORIGINAL PAPER Ordered median problem with demand distribution weights Włodzimierz Ogryczak 1 Paweł Olender 2 Received: 30 March 2015 / Accepted:

More information

Integer programming: an introduction. Alessandro Astolfi

Integer programming: an introduction. Alessandro Astolfi Integer programming: an introduction Alessandro Astolfi Outline Introduction Examples Methods for solving ILP Optimization on graphs LP problems with integer solutions Summary Introduction Integer programming

More information

Outline. 15. Descriptive Summary, Design, and Inference. Descriptive summaries. Data mining. The centroid

Outline. 15. Descriptive Summary, Design, and Inference. Descriptive summaries. Data mining. The centroid Outline 15. Descriptive Summary, Design, and Inference Geographic Information Systems and Science SECOND EDITION Paul A. Longley, Michael F. Goodchild, David J. Maguire, David W. Rhind 2005 John Wiley

More information

Place Syntax Tool (PST)

Place Syntax Tool (PST) Place Syntax Tool (PST) Alexander Ståhle To cite this report: Alexander Ståhle (2012) Place Syntax Tool (PST), in Angela Hull, Cecília Silva and Luca Bertolini (Eds.) Accessibility Instruments for Planning

More information

Multi-objective Quadratic Assignment Problem instances generator with a known optimum solution

Multi-objective Quadratic Assignment Problem instances generator with a known optimum solution Multi-objective Quadratic Assignment Problem instances generator with a known optimum solution Mădălina M. Drugan Artificial Intelligence lab, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels,

More information

A technical appendix for multihoming and compatibility

A technical appendix for multihoming and compatibility A technical appendix for multihoming and compatibility Toker Doganoglu and Julian Wright July 19, 2005 We would like to thank two anonymous referees and our editor, Simon Anderson, for their very helpful

More information

Modern Urban and Regional Economics

Modern Urban and Regional Economics Modern Urban and Regional Economics SECOND EDITION Philip McCann OXFORD UNIVERSITY PRESS Contents List of figures List of tables Introduction xii xiv xvii Part I Urban and Regional Economic Models and

More information

JOINT PRICING AND NETWORK CAPACITY SETTING PROBLEM

JOINT PRICING AND NETWORK CAPACITY SETTING PROBLEM Advanced OR and AI Methods in Transportation JOINT PRICING AND NETWORK CAPACITY SETTING PROBLEM Luce BROTCORNE, Patrice MARCOTTE, Gilles SAVARD, Mickael WIART Abstract. We consider the problem of jointly

More information

Spatial Thinking and Modeling of Network-Based Problems

Spatial Thinking and Modeling of Network-Based Problems Spatial Thinking and Modeling of Network-Based Problems Presentation at the SPACE Workshop Columbus, Ohio, July 1, 25 Shih-Lung Shaw Professor Department of Geography University of Tennessee Knoxville,

More information

Multi level inventory management decisions with transportation cost consideration in fuzzy environment. W. Ritha, S.

Multi level inventory management decisions with transportation cost consideration in fuzzy environment. W. Ritha, S. Annals of Fuzzy Mathematics and Informatics Volume 2, No. 2, October 2011, pp. 171-181 ISSN 2093 9310 http://www.afmi.or.kr @FMI c Kyung Moon Sa Co. http://www.kyungmoon.com Multi level inventory management

More information

The Leader Follower Location Model

The Leader Follower Location Model Netw Spat Econ (2007) 7:45 61 DOI 10.1007/s11067-006-9007-2 The Leader Follower Location Model D. R. Santos-Peñate R. Suárez-Vega P. Dorta-González Published online: 6 December 2006 Springer Science +

More information

Performance Evaluation of Tehran-Qom Highway Emergency Medical Service System Using Hypercube Queuing Model

Performance Evaluation of Tehran-Qom Highway Emergency Medical Service System Using Hypercube Queuing Model Western New England University From the SelectedWorks of Mohammadsadegh Mobin Summer May 2, 2015 Performance Evaluation of Tehran-Qom Highway Emergency Medical Service System Using Hypercube Queuing Model

More information

Firefly algorithm in optimization of queueing systems

Firefly algorithm in optimization of queueing systems BULLETIN OF THE POLISH ACADEMY OF SCIENCES TECHNICAL SCIENCES, Vol. 60, No. 2, 2012 DOI: 10.2478/v10175-012-0049-y VARIA Firefly algorithm in optimization of queueing systems J. KWIECIEŃ and B. FILIPOWICZ

More information

Constrained Inventory Allocation

Constrained Inventory Allocation Constrained Inventory Allocation Phillip G. Bradford Michael N. Katehakis January 25, 2006 SYNOPTIC ABSTRACT Supply constraints may include minimum supply commitments and maximum delivery limits. Given

More information

THECENTRALWAREHOUSELOCATION PROBLEM REVISITED

THECENTRALWAREHOUSELOCATION PROBLEM REVISITED THECENTRALWAREHOUSELOCATION PROBLEM REVISITED ZVI DREZNER College of Business and Economics California State University-Fullerton Fullerton, CA 92834 Tel: (714) 278-2712, e-mail: zdrezner@fullerton.edu

More information

AP Human Geography Syllabus

AP Human Geography Syllabus AP Human Geography Syllabus Textbook The Cultural Landscape: An Introduction to Human Geography. Rubenstein, James M. 10 th Edition. Upper Saddle River, N.J.: Prentice Hall 2010 Course Objectives This

More information

Key Issue 1: Where Are Services Distributed?

Key Issue 1: Where Are Services Distributed? Key Issue 1: Where Are Services Distributed? Pages 430-433 *See the Introduction on page 430 to answer questions #1-4 1. Define service: 2. What sector of the economy do services fall under? 3. Define

More information

Multi-level Facility Location as the Maximization of a Submodular Set Function

Multi-level Facility Location as the Maximization of a Submodular Set Function Multi-level Facility Location as the Maximization of a Submodular Set Function Camilo Ortiz-Astorquiza a, Ivan Contreras a, Gilbert Laporte b a Concordia University and Interuniversity Research Centre

More information

CHAPTER-3 MULTI-OBJECTIVE SUPPLY CHAIN NETWORK PROBLEM

CHAPTER-3 MULTI-OBJECTIVE SUPPLY CHAIN NETWORK PROBLEM CHAPTER-3 MULTI-OBJECTIVE SUPPLY CHAIN NETWORK PROBLEM 3.1 Introduction A supply chain consists of parties involved, directly or indirectly, in fulfilling customer s request. The supply chain includes

More information

Trip Distribution Modeling Milos N. Mladenovic Assistant Professor Department of Built Environment

Trip Distribution Modeling Milos N. Mladenovic Assistant Professor Department of Built Environment Trip Distribution Modeling Milos N. Mladenovic Assistant Professor Department of Built Environment 25.04.2017 Course Outline Forecasting overview and data management Trip generation modeling Trip distribution

More information

Key Issue 1: Where Are Services Distributed? INTRODUCING SERVICES AND SETTLEMENTS LEARNING OUTCOME DESCRIBE THE THREE TYPES OF SERVICES

Key Issue 1: Where Are Services Distributed? INTRODUCING SERVICES AND SETTLEMENTS LEARNING OUTCOME DESCRIBE THE THREE TYPES OF SERVICES Revised 2017 NAME: PERIOD: Rubenstein: The Cultural Landscape (12 th edition) Chapter Twelve Services and Settlements (pages 430 thru 457) This is the primary means by which you will be taking notes this

More information

How does the complexity of a road network affect optimal facility locations?

How does the complexity of a road network affect optimal facility locations? Working papers in transport, tourism, information technology and microdata analysis How does the complexity of a road network affect optimal facility locations? Author 1: Xiaoyun Zhao Author 2: Pascal

More information

Stochastic programs with binary distributions: Structural properties of scenario trees and algorithms

Stochastic programs with binary distributions: Structural properties of scenario trees and algorithms INSTITUTT FOR FORETAKSØKONOMI DEPARTMENT OF BUSINESS AND MANAGEMENT SCIENCE FOR 12 2017 ISSN: 1500-4066 October 2017 Discussion paper Stochastic programs with binary distributions: Structural properties

More information

Reduction of Random Variables in Structural Reliability Analysis

Reduction of Random Variables in Structural Reliability Analysis Reduction of Random Variables in Structural Reliability Analysis S. Adhikari and R. S. Langley Department of Engineering University of Cambridge Trumpington Street Cambridge CB2 1PZ (U.K.) February 21,

More information

Topic 6: Projected Dynamical Systems

Topic 6: Projected Dynamical Systems Topic 6: Projected Dynamical Systems John F. Smith Memorial Professor and Director Virtual Center for Supernetworks Isenberg School of Management University of Massachusetts Amherst, Massachusetts 01003

More information

Design Patent Damages under Sequential Innovation

Design Patent Damages under Sequential Innovation Design Patent Damages under Sequential Innovation Yongmin Chen and David Sappington University of Colorado and University of Florida February 2016 1 / 32 1. Introduction Patent policy: patent protection

More information

MODULE -4 BAYEIAN LEARNING

MODULE -4 BAYEIAN LEARNING MODULE -4 BAYEIAN LEARNING CONTENT Introduction Bayes theorem Bayes theorem and concept learning Maximum likelihood and Least Squared Error Hypothesis Maximum likelihood Hypotheses for predicting probabilities

More information

Lecture 8: Column Generation

Lecture 8: Column Generation Lecture 8: Column Generation (3 units) Outline Cutting stock problem Classical IP formulation Set covering formulation Column generation A dual perspective Vehicle routing problem 1 / 33 Cutting stock

More information

Appendix A.0: Approximating other performance measures

Appendix A.0: Approximating other performance measures Appendix A.0: Approximating other performance measures Alternative definition of service level and approximation. The waiting time is defined as the minimum of virtual waiting time and patience. Define

More information

UC Santa Barbara CSISS Classics

UC Santa Barbara CSISS Classics UC Santa Barbara CSISS Classics Title Walter Christaller, Hierarchical Patterns of Urbanization. CSISS Classics Permalink https://escholarship.org/uc/item/6188p69v Author Agarwal, Pragya Publication Date

More information

Microeconomic theory focuses on a small number of concepts. The most fundamental concept is the notion of opportunity cost.

Microeconomic theory focuses on a small number of concepts. The most fundamental concept is the notion of opportunity cost. Microeconomic theory focuses on a small number of concepts. The most fundamental concept is the notion of opportunity cost. Opportunity Cost (or "Wow, I coulda had a V8!") The underlying idea is derived

More information

Variable Objective Search

Variable Objective Search Variable Objective Search Sergiy Butenko, Oleksandra Yezerska, and Balabhaskar Balasundaram Abstract This paper introduces the variable objective search framework for combinatorial optimization. The method

More information

many authors. For an overview see e.g. [5, 6, 3, 6]. In practice the modeling of the investigated region as the complete IR is not realistic. There ma

many authors. For an overview see e.g. [5, 6, 3, 6]. In practice the modeling of the investigated region as the complete IR is not realistic. There ma Planar Weber Location Problems with Line Barriers K. Klamroth Λ Universität Kaiserslautern Abstract The Weber problem for a given finite set of existing facilities in the plane is to find the location

More information

Appendix: Simple Methods for Shift Scheduling in Multi-Skill Call Centers

Appendix: Simple Methods for Shift Scheduling in Multi-Skill Call Centers Appendix: Simple Methods for Shift Scheduling in Multi-Skill Call Centers Sandjai Bhulai, Ger Koole & Auke Pot Vrije Universiteit, De Boelelaan 1081a, 1081 HV Amsterdam, The Netherlands Supplementary Material

More information

An equity-efficiency trade-off in a geometric approach to committee selection Daniel Eckert and Christian Klamler

An equity-efficiency trade-off in a geometric approach to committee selection Daniel Eckert and Christian Klamler An equity-efficiency trade-off in a geometric approach to committee selection Daniel Eckert and Christian Klamler An equity-efficiency trade-off in a geometric approach to committee selection Daniel Eckert

More information

King Fahd University of Petroleum & Minerals College of Engineering Sciences Civil Engineering Department. Geographical Information Systems(GIS)

King Fahd University of Petroleum & Minerals College of Engineering Sciences Civil Engineering Department. Geographical Information Systems(GIS) King Fahd University of Petroleum & Minerals College of Engineering Sciences Civil Engineering Department Geographical Information Systems(GIS) Term Project Titled Delineating Potential Area for Locating

More information

Recognizing single-peaked preferences on aggregated choice data

Recognizing single-peaked preferences on aggregated choice data Recognizing single-peaked preferences on aggregated choice data Smeulders B. KBI_1427 Recognizing Single-Peaked Preferences on Aggregated Choice Data Smeulders, B. Abstract Single-Peaked preferences play

More information

Session-Based Queueing Systems

Session-Based Queueing Systems Session-Based Queueing Systems Modelling, Simulation, and Approximation Jeroen Horters Supervisor VU: Sandjai Bhulai Executive Summary Companies often offer services that require multiple steps on the

More information

Week 10: Theory of the Firm (Jehle and Reny, Chapter 3)

Week 10: Theory of the Firm (Jehle and Reny, Chapter 3) Week 10: Theory of the Firm (Jehle and Reny, Chapter 3) Tsun-Feng Chiang* *School of Economics, Henan University, Kaifeng, China November 22, 2015 First Last (shortinst) Short title November 22, 2015 1

More information

Chapter 12. Services

Chapter 12. Services Chapter 12 Services Where di services originate? Key Issue #1 Shoppers in Salzburg, Austria Origins & Types of Services Types of services Consumer services Business services Public services Changes in

More information

Assortment Optimization under the Multinomial Logit Model with Nested Consideration Sets

Assortment Optimization under the Multinomial Logit Model with Nested Consideration Sets Assortment Optimization under the Multinomial Logit Model with Nested Consideration Sets Jacob Feldman School of Operations Research and Information Engineering, Cornell University, Ithaca, New York 14853,

More information

Inequality measures and equitable approaches to location problems q

Inequality measures and equitable approaches to location problems q European Journal of Operational Research 122 (2000) 374±391 www.elsevier.com/locate/orms Inequality measures and equitable approaches to location problems q Wøodzimierz Ogryczak * Institute of Informatics,

More information

An artificial chemical reaction optimization algorithm for. multiple-choice; knapsack problem.

An artificial chemical reaction optimization algorithm for. multiple-choice; knapsack problem. An artificial chemical reaction optimization algorithm for multiple-choice knapsack problem Tung Khac Truong 1,2, Kenli Li 1, Yuming Xu 1, Aijia Ouyang 1, and Xiaoyong Tang 1 1 College of Information Science

More information

The National Spatial Strategy

The National Spatial Strategy Purpose of this Consultation Paper This paper seeks the views of a wide range of bodies, interests and members of the public on the issues which the National Spatial Strategy should address. These views

More information

University of Groningen. The binary knapsack problem Ghosh, Diptesh; Goldengorin, Boris

University of Groningen. The binary knapsack problem Ghosh, Diptesh; Goldengorin, Boris University of Groningen The binary knapsack problem Ghosh, Diptesh; Goldengorin, Boris IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it.

More information

Application for Geotourism Charter

Application for Geotourism Charter Application for Geotourism Charter Thank you for the interest in geotourism as a long-term strategy to foster wisely managed tourism and enlightened destination stewardship. National Geographic s Center

More information

Linear & nonlinear classifiers

Linear & nonlinear classifiers Linear & nonlinear classifiers Machine Learning Hamid Beigy Sharif University of Technology Fall 1396 Hamid Beigy (Sharif University of Technology) Linear & nonlinear classifiers Fall 1396 1 / 44 Table

More information

A route map to calibrate spatial interaction models from GPS movement data

A route map to calibrate spatial interaction models from GPS movement data A route map to calibrate spatial interaction models from GPS movement data K. Sila-Nowicka 1, A.S. Fotheringham 2 1 Urban Big Data Centre School of Political and Social Sciences University of Glasgow Lilybank

More information

Formulas to Estimate the VRP Average Distance Traveled in Elliptic Zones

Formulas to Estimate the VRP Average Distance Traveled in Elliptic Zones F. Robuste, M. Estrada & A. Lopez-Pita 1 Formulas to Estimate the VRP Average Distance Traveled in Elliptic Zones F. Robuste, M. Estrada & A. Lopez Pita CENIT Center for Innovation in Transport Technical

More information

A new nonmonotone Newton s modification for unconstrained Optimization

A new nonmonotone Newton s modification for unconstrained Optimization A new nonmonotone Newton s modification for unconstrained Optimization Aristotelis E. Kostopoulos a George S. Androulakis b a a Department of Mathematics, University of Patras, GR-265.04, Rio, Greece b

More information

National Center for Geographic Information and Analysis. A Comparison of Strategies for Data Storage Reduction in Location-Allocation Problems

National Center for Geographic Information and Analysis. A Comparison of Strategies for Data Storage Reduction in Location-Allocation Problems National Center for Geographic Information and Analysis A Comparison of Strategies for Data Storage Reduction in Location-Allocation Problems by Paul A. Sorensen Richard L. Church University of California,

More information

Research Article A Novel Differential Evolution Invasive Weed Optimization Algorithm for Solving Nonlinear Equations Systems

Research Article A Novel Differential Evolution Invasive Weed Optimization Algorithm for Solving Nonlinear Equations Systems Journal of Applied Mathematics Volume 2013, Article ID 757391, 18 pages http://dx.doi.org/10.1155/2013/757391 Research Article A Novel Differential Evolution Invasive Weed Optimization for Solving Nonlinear

More information

Fundamental Theorems of Welfare Economics

Fundamental Theorems of Welfare Economics Fundamental Theorems of Welfare Economics Ram Singh Lecture 6 September 29, 2015 Ram Singh: (DSE) General Equilibrium Analysis September 29, 2015 1 / 14 First Fundamental Theorem The First Fundamental

More information

How does data quality in a network affect heuristic

How does data quality in a network affect heuristic Working papers in transport, tourism, information technology and microdata analysis How does data quality in a network affect heuristic solutions? Ev. underrubrik ska stå i Arial, 14pt, Fet stil Mengjie

More information

Network design for a service operation with lost demand and possible disruptions

Network design for a service operation with lost demand and possible disruptions Network design for a service operation with lost demand and possible disruptions Opher Baron, Oded Berman, Yael Deutsch Joseph L. Rotman School of Management, University of Toronto, 105 St. George St.,

More information

where u is the decision-maker s payoff function over her actions and S is the set of her feasible actions.

where u is the decision-maker s payoff function over her actions and S is the set of her feasible actions. Seminars on Mathematics for Economics and Finance Topic 3: Optimization - interior optima 1 Session: 11-12 Aug 2015 (Thu/Fri) 10:00am 1:00pm I. Optimization: introduction Decision-makers (e.g. consumers,

More information

Microeconomic Theory. Microeconomic Theory. Everyday Economics. The Course:

Microeconomic Theory. Microeconomic Theory. Everyday Economics. The Course: The Course: Microeconomic Theory This is the first rigorous course in microeconomic theory This is a course on economic methodology. The main goal is to teach analytical tools that will be useful in other

More information

OPTIMALITY OF RANDOMIZED TRUNK RESERVATION FOR A PROBLEM WITH MULTIPLE CONSTRAINTS

OPTIMALITY OF RANDOMIZED TRUNK RESERVATION FOR A PROBLEM WITH MULTIPLE CONSTRAINTS OPTIMALITY OF RANDOMIZED TRUNK RESERVATION FOR A PROBLEM WITH MULTIPLE CONSTRAINTS Xiaofei Fan-Orzechowski Department of Applied Mathematics and Statistics State University of New York at Stony Brook Stony

More information

Tradable Permits for System-Optimized Networks. Anna Nagurney Isenberg School of Management University of Massachusetts Amherst, MA 01003

Tradable Permits for System-Optimized Networks. Anna Nagurney Isenberg School of Management University of Massachusetts Amherst, MA 01003 Tradable Permits for System-Optimized Networks Anna Nagurney Isenberg School of Management University of Massachusetts Amherst, MA 01003 c 2002 Introduction In this lecture, I return to the policy mechanism

More information

Optimizing departure times in vehicle routes

Optimizing departure times in vehicle routes Optimizing departure times in vehicle routes A.L. Kok, E.W. Hans, J.M.J. Schutten Operational Methods for Production and Logistics, University of Twente, P.O. Box 217, 7500AE, Enschede, Netherlands Abstract

More information

A Modified Gravity p-median Model for Optimizing Facility Locations

A Modified Gravity p-median Model for Optimizing Facility Locations Journal of Systems Science and Information Oct., 2018, Vol. 6, No. 5, pp. 421 434 DOI: 10.21078/JSSI-2018-421-14 A Modified Gravity p-median Model for Optimizing Facility Locations Zhuolin TAO College

More information

C31: Game Theory, Lecture 1

C31: Game Theory, Lecture 1 C31: Game Theory, Lecture 1 V. Bhaskar University College London 5 October 2006 C31 Lecture 1: Games in strategic form & Pure strategy equilibrium Osborne: ch 2,3, 12.2, 12.3 A game is a situation where:

More information

SPATIAL ANALYSIS. Transformation. Cartogram Central. 14 & 15. Query, Measurement, Transformation, Descriptive Summary, Design, and Inference

SPATIAL ANALYSIS. Transformation. Cartogram Central. 14 & 15. Query, Measurement, Transformation, Descriptive Summary, Design, and Inference 14 & 15. Query, Measurement, Transformation, Descriptive Summary, Design, and Inference Geographic Information Systems and Science SECOND EDITION Paul A. Longley, Michael F. Goodchild, David J. Maguire,

More information

Information Choice in Macroeconomics and Finance.

Information Choice in Macroeconomics and Finance. Information Choice in Macroeconomics and Finance. Laura Veldkamp New York University, Stern School of Business, CEPR and NBER Spring 2009 1 Veldkamp What information consumes is rather obvious: It consumes

More information

Optimization Problems with Probabilistic Constraints

Optimization Problems with Probabilistic Constraints Optimization Problems with Probabilistic Constraints R. Henrion Weierstrass Institute Berlin 10 th International Conference on Stochastic Programming University of Arizona, Tucson Recommended Reading A.

More information

A BRANCH&BOUND ALGORITHM FOR SOLVING ONE-DIMENSIONAL CUTTING STOCK PROBLEMS EXACTLY

A BRANCH&BOUND ALGORITHM FOR SOLVING ONE-DIMENSIONAL CUTTING STOCK PROBLEMS EXACTLY APPLICATIONES MATHEMATICAE 23,2 (1995), pp. 151 167 G. SCHEITHAUER and J. TERNO (Dresden) A BRANCH&BOUND ALGORITHM FOR SOLVING ONE-DIMENSIONAL CUTTING STOCK PROBLEMS EXACTLY Abstract. Many numerical computations

More information

Network Flows. 6. Lagrangian Relaxation. Programming. Fall 2010 Instructor: Dr. Masoud Yaghini

Network Flows. 6. Lagrangian Relaxation. Programming. Fall 2010 Instructor: Dr. Masoud Yaghini In the name of God Network Flows 6. Lagrangian Relaxation 6.3 Lagrangian Relaxation and Integer Programming Fall 2010 Instructor: Dr. Masoud Yaghini Integer Programming Outline Branch-and-Bound Technique

More information

12 Discriminant Analysis

12 Discriminant Analysis 12 Discriminant Analysis Discriminant analysis is used in situations where the clusters are known a priori. The aim of discriminant analysis is to classify an observation, or several observations, into

More information

CHAPTER 2: QUADRATIC PROGRAMMING

CHAPTER 2: QUADRATIC PROGRAMMING CHAPTER 2: QUADRATIC PROGRAMMING Overview Quadratic programming (QP) problems are characterized by objective functions that are quadratic in the design variables, and linear constraints. In this sense,

More information

Distance Metrics and Fitness Distance Analysis for the Capacitated Vehicle Routing Problem

Distance Metrics and Fitness Distance Analysis for the Capacitated Vehicle Routing Problem MIC2005. The 6th Metaheuristics International Conference 603 Metrics and Analysis for the Capacitated Vehicle Routing Problem Marek Kubiak Institute of Computing Science, Poznan University of Technology

More information

A CHARACTERIZATION OF STRICT LOCAL MINIMIZERS OF ORDER ONE FOR STATIC MINMAX PROBLEMS IN THE PARAMETRIC CONSTRAINT CASE

A CHARACTERIZATION OF STRICT LOCAL MINIMIZERS OF ORDER ONE FOR STATIC MINMAX PROBLEMS IN THE PARAMETRIC CONSTRAINT CASE Journal of Applied Analysis Vol. 6, No. 1 (2000), pp. 139 148 A CHARACTERIZATION OF STRICT LOCAL MINIMIZERS OF ORDER ONE FOR STATIC MINMAX PROBLEMS IN THE PARAMETRIC CONSTRAINT CASE A. W. A. TAHA Received

More information

Page No. (and line no. if applicable):

Page No. (and line no. if applicable): COALITION/IEC (DAYMARK LOAD) - 1 COALITION/IEC (DAYMARK LOAD) 1 Tab and Daymark Load Forecast Page No. Page 3 Appendix: Review (and line no. if applicable): Topic: Price elasticity Sub Topic: Issue: Accuracy

More information

Algebraic Equations. 2.0 Introduction. Nonsingular versus Singular Sets of Equations. A set of linear algebraic equations looks like this:

Algebraic Equations. 2.0 Introduction. Nonsingular versus Singular Sets of Equations. A set of linear algebraic equations looks like this: Chapter 2. 2.0 Introduction Solution of Linear Algebraic Equations A set of linear algebraic equations looks like this: a 11 x 1 + a 12 x 2 + a 13 x 3 + +a 1N x N =b 1 a 21 x 1 + a 22 x 2 + a 23 x 3 +

More information

CIV3703 Transport Engineering. Module 2 Transport Modelling

CIV3703 Transport Engineering. Module 2 Transport Modelling CIV3703 Transport Engineering Module Transport Modelling Objectives Upon successful completion of this module you should be able to: carry out trip generation calculations using linear regression and category

More information

Logistics. Lecture notes. Maria Grazia Scutellà. Dipartimento di Informatica Università di Pisa. September 2015

Logistics. Lecture notes. Maria Grazia Scutellà. Dipartimento di Informatica Università di Pisa. September 2015 Logistics Lecture notes Maria Grazia Scutellà Dipartimento di Informatica Università di Pisa September 2015 These notes are related to the course of Logistics held by the author at the University of Pisa.

More information

CIE4801 Transportation and spatial modelling Modal split

CIE4801 Transportation and spatial modelling Modal split CIE4801 Transportation and spatial modelling Modal split Rob van Nes, Transport & Planning 31-08-18 Delft University of Technology Challenge the future Content Nested logit part 2 Modelling component 3:

More information

Information theoretic solutions for correlated bivariate processes

Information theoretic solutions for correlated bivariate processes Economics Letters 97 (2007) 201 207 www.elsevier.com/locate/econbase Information theoretic solutions for correlated bivariate processes Wendy K. Tam Cho a,, George G. Judge b a Departments of Political

More information

The central limit theorem

The central limit theorem 14 The central limit theorem The central limit theorem is a refinement of the law of large numbers For a large number of independent identically distributed random variables X 1,,X n, with finite variance,

More information