Papers by M. Angelica Salazar Aguilar
Computación y Sistemas, Sep 1, 2006
Time series forecasting is an active area for the application of Artificial Neural Networks (ANNs... more Time series forecasting is an active area for the application of Artificial Neural Networks (ANNs). Although the selection of an ANN has been greatly simplified, it remains a challenge to adequately determine the ANN's parameters. In this work a method based on statistical analysis and optimization techniques is proposed to select the ANN's parameters for application in time series forecasting. The results on the successful application of the method in a real demand forecasting problem for the telecommunications industry are ...
Commercial Territory Design for a Distribution Firm with New Constructive and Destructive Heuristics
A commercial territory design problem with compactness maximization criterion subject to territor... more A commercial territory design problem with compactness maximization criterion subject to territory balancing and connectivity is addressed. Four new heuristics based on Greedy Randomized Adaptive Search Procedures within a location-allocation scheme for this NP-hard combinatorial optimization problem are proposed. The first three (named GRLH1, GRLH2, and GRDL) build the territories simultaneously. Their construction phase consists of two parts: a location
The multi-district team orienteering problem
Computers & Operations Research, 2014
ABSTRACT This paper introduces the multi-district team orienteering problem. In this problem, one... more ABSTRACT This paper introduces the multi-district team orienteering problem. In this problem, one must schedule a set of mandatory and optional tasks located in several districts, within a planning horizon. The total available time determined by the length of the planning horizon must be distributed among the districts. All mandatory tasks within each district must be performed, while the other tasks can be performed if time allows. A positive profit or score is collected whenever an optional task is performed. Additionally, some incompatibility constraints between tasks are taken into account. The objective is to determine a schedule for a set of tasks to be performed daily within each district, while maximizing the total collected profit. A mixed integer formulation and an adaptive large neighborhood search heuristic are proposed for this problem. The performance of the proposed algorithm is assessed over a large set of randomly generated instances. Computational results confirm the efficiency of the algorithm.
Statistical Characterization and Optimization of Artificial Neural Networks in Time Series Forecasting: The One-Period Forecast Case
Computación y …, Jan 1, 2006
Synchronized Arc Routing for Snow Plowing Operations
This paper introduces a synchronized arc routing problem for snow plowing operations. In this pro... more This paper introduces a synchronized arc routing problem for snow plowing operations. In this problem, routes must be designed in such a way that street segments with two or more lanes in the same direction are plowed simultaneously by different synchronized vehicles. A mixed integer formulation and an adaptive large neighborhood search heuristic are proposed. The performance of the proposed algorithm is evaluated over a large instance set, including artificial and real data. Computational results confirm the efficiency of the algorithm.
A bi-objective programming model for designing compact and balanced territories in commercial districting
… Research Part C: …, Jan 1, 2010
In this paper, we address a territory design problem arising from a bottled beverage distribution... more In this paper, we address a territory design problem arising from a bottled beverage distribution company. We propose a bi-objective programming model where dispersion and balancing with respect to the number of customers are used as performance criteria. Constraints such as connectivity and balancing with respect to sales volume are considered in the model. Most of the work in territory design has been developed for single-objective models. To the best of our knowledge, this is the first multi-objective approach for this commercial territory design problem, and in particular, for territory design with connectivity constraints. We propose an improved ε-constraint method for generating the optimal Pareto front. Empirical evidence over a variety of instances shows that the improved method is well suited for finding optimal Pareto fronts with no more computational effort than the traditional method. Instances of up to 150 units and 6 territories are solved in relatively short amount of time. For this problem, the improved method finds practically the same fronts than those found by the traditional ε-constraint method. In addition, we observe that when the firm reduces the tolerance in the imbalance of sales volume the efficient fronts change and when the number of territories increases, the balance with respect to the number of customers becomes harder to achieve.
GRASP strategies for a bi-objective commercial territory design problem
Journal of Heuristics, Jan 1, 2011
A bi-objective commercial territory design problem motivated by a real-world application from the... more A bi-objective commercial territory design problem motivated by a real-world application from the bottled beverage distribution industry is addressed. The problem considers territory compactness and balancing with respect to number of customers as optimization criteria. Previous work has focused on exact methods for small- to medium-scale instances. In this work, a GRASP framework is proposed for tackling considerably large instances. Within this framework two general schemes are developed. For each of these schemes two strategies are studied: (i) keeping connectivity as a hard constraint during construction and post-processing phases and, (ii) ignoring connectivity during the construction phase and adding this as another minimizing objective function during the post-processing phase. These strategies are empirically evaluated and compared to NSGA-II, one of the most successful evolutionary methods known in literature. Computational results show the superiority of the proposed strategies. In addition, one of the proposed GRASP strategies is successfully applied to a case study from industry.
Networks and Spatial …, Jan 1, 2011
In this work, a series of novel formulations for a commercial territory design problem motivated ... more In this work, a series of novel formulations for a commercial territory design problem motivated by a real-world case are proposed. The problem consists on determining a partition of a set of units located in a territory that meets multiple criteria such as compactness, connectivity, and balance in terms of customers and product demand. Thus far, different versions of this problem have been approached with heuristics due to its NP-completeness. The proposed formulations are integer quadratic programming models that involve a smaller number of variables than heretofore required. These models have also enabled the development of an exact solution framework, the first ever derived for this problem, that is based on branch and bound and a cut generation strategy. The proposed method is empirically evaluated using several instances of the new quadratic models as well as of the existing linear models. The results show that the quadratic models allow solving larger instances than the linear counterparts. The former were also observed to require fewer iterations of the exact method to converge. Based on these results the combination of the quadratic formulation and the exact method are recommended to approach problem instances associated with medium-sized cities.
Keywords Mixed-integer linear programming – Integer quadratic programming – Territory design – Location – Valid inequalities
Ingenierías, Jan 1, 2004
En este artículo se presenta una heurística (método de aproximación) que permite encontrar una se... more En este artículo se presenta una heurística (método de aproximación) que permite encontrar una secuencia de «n» tareas en un ambiente de líneas de flujo de «m» máquinas para el problema de minimizar el número de tareas que se entregan con retraso. Este procedimiento, basado en el algoritmo de Moore para problemas de una máquina, se evalúa y se compara computacionalmente contra un método que genera secuencias, sin tomar en cuenta la estructura del problema, en una variedad de problemas bajo dos escenarios distintos: fechas de entrega estrictas y no estrictas. Se observa que el procedimiento propuesto brinda mejores resultados en cuanto a la calidad de la solución encontrada.
Uploads
Papers by M. Angelica Salazar Aguilar
Keywords Mixed-integer linear programming – Integer quadratic programming – Territory design – Location – Valid inequalities
Keywords Mixed-integer linear programming – Integer quadratic programming – Territory design – Location – Valid inequalities