Em uma cadeia produtiva de minério, as operações de pátio de estocagem, principalmente as de recuperação de pilhas de minério, exercem um papel fundamental por vincularem os processos de beneficiamento e de transporte. Com esta finalidade, esta dissertação propõe o sequenciamento das pilhas de minério a serem recuperadas através da adaptação de heurísticas trazidas pela literatura. Primeiramente, propõe-se uma heurística de sequenciamento para a retomada de pilhas de minério (entendidas como tarefas a serem sequenciadas) nos seus respectivos pátios através da aplicação de um índice de priorização de pilhas (IP). Tal índice apoia-se em fatores relevantes para as operações de pátios de estocagem, como capacidade das recuperadoras, qualidade e tempo de residência do minério e tempo de deslocamento entre pilhas (setup). A segunda heurística proposta, ATCSM (Apparent Tardiness Cost with Setups for mineral recovery), modifica a regra de despacho ATCS (Apparent Tardiness Cost with Setups) com vistas à sua aplicação na retomada de pilhas de minério em pátios de recuperação. O ATCSM apoia-se em fatores tidos como relevantes para as operações de pátios de estocagem, como tempo disponível para o empilhamento de um produto e data de entrega de uma pilha, entre outros. Os métodos propostos foram aplicados em um sistema de recuperação de minério composto por dois pátios, duas máquinas recuperadoras e doze pilhas. As sequências de recuperação geradas pelas heurísticas propostas foram consideradas coerentes por especialistas de pátio de estocagem de uma empresa mineradora. / In the supply chain of ore operations, stockyard operations, mainly the recovery of ore piles, play a fundamental role between beneficiation and transport processes. This thesis proposes new scheduling heuristics for sequencing ore piles recovering order. The first heuristics proposes a pile prioritization index (IP) that relies on relevant factors for stockyards operations, including machines capability, ore quality and residence time, and travel time between piles (setup). With similar purposes, the second heuristics modifies the dispatching rule ATCS (Apparent Tardiness Cost with Setups), yielding the ATCSM (Apparent Tardiness Cost with Setups for mineral recovery). The proposed ATCSM also relies on factors regarded as relevant to stockyard operations, as time available for stacking of a product and piles due date, among others. The proposed methods were applied to an ore recovery system composed of two stockyards, two recovery machines and twelve piles. The recovery sequences generated by both heuristics were considered consistent by experts from a mining company.

As school choice opportunities have become more prevalent and information about schools more readily available, there is still a lack of understanding of how parents use information to evaluate schools. The discussion around school judgment-making predominately focuses on whether parents know about school choice and the constraints parents face which limit choice, but I investigate, using 91 interviews of parents living in a low socio-economic community, how parents make judgments and evaluate schools past the discussion of what schools are available to parents and the constraints those parents face. The results of this study are that parents use heuristics—specifically familiarity, endorsement, and representativeness—to help them make judgments about schools. Knowing that parents use heuristics, policy-makers and educators can better address these parents needs and provide information that is more beneficial to them for making judgments about schools.

The quadratic assignment problem (QAP) is known to be one of the most computationally difficult combinatorial problems. Optimally solvable instances of the QAP remain of size n ≤ 40 with heuristics used to solve instances in the range 40 ≤ n ≤ 256. In this thesis we develop a local optimization algorithm called GradSwaps (GS). GS uses the first-order Taylor approximation (FOA) to efficiently determine improving swaps in the solution. We use GS to locally optimize instances of the QAP of size 1000 ≤ n ≤ 70000 where the data matrices are given in factored form, enabling efficient computations. We give theoretical background and justification for using the FOA and bound the error inherent in the approximation. A strategy for extending GS to larger scale QAPs using blocks of indices is described in detail. Three novel large-scale applications of the QAP are developed. First, a strategy for data visualization using an extreme learning machine (ELM) where the quality of the visualization is measured in the original data space instead of the projected space. Second, a version of the traveling salesperson problem (TSP) with the squared Euclidean distance metric; this distance metric allows the factorization of the data matrix, a key component for using GS. Third, a method for generating random data with designated distribution and correlation to an accuracy surpassing traditional techniques.

This thesis explores through the creation of four artworks, the nature of the sublime ruin. To facilitate this it examines the disintegration of selected religious feminine metaphors. The artworks are rendered through a multiplication of layers bound by translucent/transparent resin. These are produced to examine the potentials between traditional craft and contemporary digital mediums, thus creating sites where eroding fragments may express an excess of meaning through enigmatic construction.

We consider the survivable network design problem - the problem of designing, at minimum cost, a network with edge-connectivity requirements. As special cases, this problem encompasses the Steiner tree problem, the traveling salesman problem and the k-connected network design problem. We establish a property, referred to as the parsimonious property, of the linear programming (LP) relaxation of a classical formulation for the problem. The parsimonious property has numerous consequences. For example, we derive various structural properties of these LP relaxations, we present some algorithmic improvements and we perform tight worstcase analyses of two heuristics for the survivable network design problem.

Individuals utilize heuristics in order to simplify problems, which may lead to biases in decision-making. The research question of this study is: “How does the affect heuristic impact the investment process of private equity decision-makers reviewing proposals?” Through an exploratory multi-case analysis, insight is provided into complex private equity decisions by studying biases in the investment process. This is a study of private equity groups’ (PEG) decision-making process when they consider businesses for investment. Qualitative data was generated from semi-structured interviews with twenty private equity decision-makers. The deliberative heuristics applied in the teaser review are learned from process experience and guide the deliberation on whether to proceed. Simplifying heuristics are applied in the more informal review process. Organizational learning was exhibited as the PEGs have modified their investment structures based on previous experiences. The study indicates that experience and learning lead to the construction of an affect heuristic that subsequently impacts investments. It also confirms the need for strategic decision-makers to recognize their own biases and adjust their processes accordingly. A significant practical implication of this study is the insight provided into the views of the PEG decision-makers as they anticipate the need to supplement the management team is helpful to business owners and their advisors. The study highlights the opportunities for biases in PEG decision-making processes. Accessing decision-makers at larger PEGs and approaching more middle market firms would broaden the results.

The development of modern pipelined and multiple functional unit processors has increased the available instruction level parallelism. In order to fully utilize these resources, compiler writers spend large amounts of time developing complex scheduling heuristics for each new architecture. In order to reduce the time spent on this process, automated machine learning techniques have been proposed to generate scheduling heuristics. We present two case studies using these techniques to generate instruction scheduling heuristics for basic blocks and super blocks. A basic block is a block of code with a single flow of control and a super block is a collection of basic blocks with a single entry point but multiple exit points. We improve previous techniques for automated generation of basic block scheduling heuristics by increasing the quality of the training data and increasing the number of features considered, including several novel features that have useful effects on scheduling instructions. Our case study into super block scheduling heuristics is a novel contribution as previous approaches were only applied to basic blocks. We show through experimentation that we can produce efficient heuristics that perform better than current heuristic methods for basic block and super block scheduling. We show that we can reduce the number of non-optimally scheduled blocks by up to 55% for basic blocks and 38% for super blocks. We also show that we can produce better schedules 7. 8 times more often than the next best heuristic for basic blocks and 4. 4 times more often for super blocks.

Less-Than-Truckload (LTL) carriers generally serve geographical regions that are more localized than the inter-city routes served by truckload carriers. That localization can lead to urban freight transportation routes that overlap. If trucks are traveling with less than full loads there may exist opportunities for carriers to collaborate over such routes. That is, Carrier A will also deliver one or more shipments of Carrier B. This will improve vehicle asset utilization and reduce asset-repositioning costs, and may also lead to reduced congestion and pollution in cities. We refer to the above coordination as “collaborative routing”. In our framework for collaboration, we also propose that carriers exchange goods at logistics platforms located at the entry point to a city. This is referred to as “entry-point collaboration”. One difficulty in collaboration is the lack of facilities to allow transfer of goods between carriers. We highlight that the reduction in pollution and congestion under our proposed framework will give the city government an incentive to support these initiatives by providing facilities. Further, our analysis has shown that contrary to the poor benefits reported by previous work on vehicle routing with transshipment, strategic location of transshipment facilities in urban areas may solve this problem and lead to large cost savings from transfer of loads between carriers. We also present a novel integrated three-phase solution method. Our first phase uses either a modified tabu search, or a guided local search, to solve the vehicle routing problems with time windows that result from entry-point collaboration. The preceding methods use a constraint programming engine for feasibility checks. The second phase uses a quad-tree search to locate facilities. Quad-tree search methods are popular in computer graphics, and for grid generation in fluid simulation. These methods are known to be efficient in partitioning a two-dimensional space for storage and computation. We use this efficiency to search a two-dimensional region and locate possible transshipment facilities. In phase three, we employ an integrated greedy local search method to build collaborative routes, using three new transshipment-specific moves for neighborhood definition. We utilize an optimization module within local search to combine multiple moves at each iteration, thereby taking efficient advantage of information from neighborhood exploration. Extensive computational tests are done on random data sets which represent a city such as Toronto. Sensitivity analysis is performed on important parameters to characterize the situations when collaboration will be beneficial. Overall results show that our proposal for collaboration leads to 12% and 15% decrease in route distance and time, respectively. Average asset utilization is seen to increase by about 5% as well.

The development of modern pipelined and multiple functional unit processors has increased the available instruction level parallelism. In order to fully utilize these resources, compiler writers spend large amounts of time developing complex scheduling heuristics for each new architecture. In order to reduce the time spent on this process, automated machine learning techniques have been proposed to generate scheduling heuristics. We present two case studies using these techniques to generate instruction scheduling heuristics for basic blocks and super blocks. A basic block is a block of code with a single flow of control and a super block is a collection of basic blocks with a single entry point but multiple exit points. We improve previous techniques for automated generation of basic block scheduling heuristics by increasing the quality of the training data and increasing the number of features considered, including several novel features that have useful effects on scheduling instructions. Our case study into super block scheduling heuristics is a novel contribution as previous approaches were only applied to basic blocks. We show through experimentation that we can produce efficient heuristics that perform better than current heuristic methods for basic block and super block scheduling. We show that we can reduce the number of non-optimally scheduled blocks by up to 55% for basic blocks and 38% for super blocks. We also show that we can produce better schedules 7. 8 times more often than the next best heuristic for basic blocks and 4. 4 times more often for super blocks.

Less-Than-Truckload (LTL) carriers generally serve geographical regions that are more localized than the inter-city routes served by truckload carriers. That localization can lead to urban freight transportation routes that overlap. If trucks are traveling with less than full loads there may exist opportunities for carriers to collaborate over such routes. That is, Carrier A will also deliver one or more shipments of Carrier B. This will improve vehicle asset utilization and reduce asset-repositioning costs, and may also lead to reduced congestion and pollution in cities. We refer to the above coordination as “collaborative routing”. In our framework for collaboration, we also propose that carriers exchange goods at logistics platforms located at the entry point to a city. This is referred to as “entry-point collaboration”. One difficulty in collaboration is the lack of facilities to allow transfer of goods between carriers. We highlight that the reduction in pollution and congestion under our proposed framework will give the city government an incentive to support these initiatives by providing facilities. Further, our analysis has shown that contrary to the poor benefits reported by previous work on vehicle routing with transshipment, strategic location of transshipment facilities in urban areas may solve this problem and lead to large cost savings from transfer of loads between carriers. We also present a novel integrated three-phase solution method. Our first phase uses either a modified tabu search, or a guided local search, to solve the vehicle routing problems with time windows that result from entry-point collaboration. The preceding methods use a constraint programming engine for feasibility checks. The second phase uses a quad-tree search to locate facilities. Quad-tree search methods are popular in computer graphics, and for grid generation in fluid simulation. These methods are known to be efficient in partitioning a two-dimensional space for storage and computation. We use this efficiency to search a two-dimensional region and locate possible transshipment facilities. In phase three, we employ an integrated greedy local search method to build collaborative routes, using three new transshipment-specific moves for neighborhood definition. We utilize an optimization module within local search to combine multiple moves at each iteration, thereby taking efficient advantage of information from neighborhood exploration. Extensive computational tests are done on random data sets which represent a city such as Toronto. Sensitivity analysis is performed on important parameters to characterize the situations when collaboration will be beneficial. Overall results show that our proposal for collaboration leads to 12% and 15% decrease in route distance and time, respectively. Average asset utilization is seen to increase by about 5% as well.