Indiana University-Purdue University Indianapolis (IUPUI) / While optimization techniques have been studied since 300 B.C. when Euclid first considered the minimal distance between a point and a line, it wasn’t until 1966 that location optimization was first applied to a problem in healthcare. Location optimization techniques are capable of increasing efficiency and equity in the placement of many types of services, including those within the healthcare industry, thus enhancing quality of life. Medical imaging is a healthcare service which helps to determine medical diagnoses in acute and preventive care settings. It provides physicians with information guiding treatment and returning a patient back to optimal health. In this study, a retrospective analysis of the locations of current medical imaging centers in central Indiana is performed, and alternate placement as determined using optimization techniques is considered and compared. This study focuses on reducing the drive time experienced by the population within the study area to their nearest imaging facility. Location optimization models such as the P-Median model, the Maximum Covering model, and Clustering and Partitioning are often used in the field of operations research to solve location problems, but are lesser known within the discipline of Geographic Information Science. This study was intended to demonstrate the capabilities of these powerful algorithms and to increase understanding of how they may be applied to problems within healthcare. While the P-Median model is effective at reducing the overall drive time for a given network set, individuals within the network may experience lengthy drive times. The results further indicate that while the Maximum Covering model is more equitable than the P-Median model, it produces large sets of assigned individuals overwhelming the capacity of one imaging center. Finally, the Clustering and Partitioning method is effective at limiting the number of individuals assigned to a given imaging center, but it does not provide information regarding average drive time for those individuals. In the end, it is determined that a capacitated Maximal Covering model would be the preferred method for solving this particular location problem.

[pt] As técnicas de agrupamentos aplicadas a um grande número de cenários de incerteza permitem a escolha de um conjunto reduzido, porém, representativo da população de cenários completa. Em outras palavras, selecionar uma amostra que contenha uma quantidade menor de elementos a ponto de reduzir suficientemente o volume total de dados e obter ganhos significativos de eficiência no processamento dos dados. Esta amostra deve, sobretudo, conseguir preservar as características do processo estocástico que o originou. Com este intuito, o presente trabalho propõe uma metodologia de seleção de cenários estocásticos utilizando o modelo clássico de Cobertura de Conjuntos, inspirada no método forward selection proposto por Heitsch e Romisch (2003). Aplicada na etapa de cálculo de demanda estocástica de ferramentas e serviços para construção de poços marítimos de exploração de petróleo, esta abordagem apresenta uma concepção de cenário diferente da usada pelos autores. O conjunto de cenários consiste em cronogramas de atividades gerados a partir da introdução de incertezas no planejamento de cada atividade, sendo eles estáticos, independentes e com múltiplos atributos. Uma análise de sensibilidade compara os resultados das demandas calculadas com os cenários selecionados pelo Problema de Cobertura de Conjuntos (PCC) e a demanda calculada com o conjunto universo de cenários. O PCC foi solucionado, nesta aplicação, em sua versão clássica da literatura a partir de um algoritmo exato e um heurístico. Os resultados apontam diferenças pouco representativas no resultado final das demandas calculadas com cenários reduzidos e com o total de cenários. A heurística, ainda que seja first solution, apresentou um resultado satisfatório em relação ao ganho de desempenho versus confiabilidade, e indica o potencial do método se aplicado em conjunto com algoritmos de metaheurística e busca local. / [en] Clustering techniques applied to a large number of scenarios under uncertainty allows the selection of a reduced, however, representative set of the complete set of scenarios. In other words, it allows to select a sample that contains a smaller amount of elements to the point of sufficiently reducing the total data volume and obtaining efficiency gains in data processing. The challenge is that the sample must, above all, be able to preserve the characteristics of the stochastic process that originated it. To this end, this study proposes a methodology for selecting stochastic scenarios using the classic Set Covering model, inspired by the forward selection method proposed by Heitsch and Romisch (2003). Applied in the calculating of stochastic demand for tools and services for the construction of offshore oil exploration wells, this approach presents a different scenario conception from the one used by the authors. The set of scenarios consists of activity schedules generated from the introduction of uncertainties in the planning of each activity, which are static, independent and with multiple attributes. A sensitivity analysis compares the results of the demands calculated with the scenarios selected by the Set Covering Problem (SCP) and the demand calculated with all the universe of scenarios. The SCP was solved, in this application, in its classic version using an exact algorithm and a heuristic algorithm. The results appoint na unexpressive loss in the final result of the demand calculated with reduced scenarios and with the complete set of scenarios. The simple first solution heuristic presented a satisfactory result in relation to the performance gain versus reliability, and indicates the potential of the method if solved with metaheuristic and local search algorithms.