An algorithm to median shortest path problem (MSPP) in the design of urban transportation networks

This paper proposes an efficient solution algorithm for realistic multi-objective median shortest path problems in the design of urban transportation networks. The proposed problem formulation and solution algorithm to median shortest path problem is based on three realistic objectives via route cost or investment cost, overall travel time of the entire network and total toll revenue. The proposed solution approach to the problem is based on the heuristic labeling and exhaustive search technique in criteria space and solution space of the algorithm respectively. The first labels each node in terms of route cost and deletes cyclic and infeasible paths in criteria space imposing cyclic break and route cost constraint respectively. The latter deletes dominated paths in terms of objectives vector in solution space in order to identify a set of Pareto optimal paths. The approach, thus, proposes a non-inferior solution set of Pareto optimal paths based on non-dominated objective vector and leaves the ultimate decision to decision-makers for purpose specific final decision during applications. A numerical experiment is conducted to test the proposed algorithm using artificial transportation network. Sensitivity analyses have shown that the proposed algorithm is advantageous and efficient over existing algorithms to find a set of Pareto optimal paths to median shortest paths problems.

M-ary trees for combinatorial asset management decision problems

A novel m-ary tree based approach is presented to solve asset management decisions which are combinatorial in nature. The approach introduces a new dynamic constraint based control mechanism which is capable of excluding infeasible solutions from the solution space. The approach also provides a solution to the challenges with ordering of assets decisions.

The deep ocean density structure at the Last Glacial Maximum : What was it and why?

The search for reliable proxies of past deep ocean temperature and salinity has proved difficult, thereby limiting our ability to understand the coupling of ocean circulation and climate over glacial-interglacial timescales. Previous inferences of deep ocean temperature and salinity from sediment pore fluid oxygen isotopes and chlorinity indicate that the deep ocean density structure at the Last Glacial Maximum (LGM, approximately 20,000 years BP) was set by salinity, and that the density contrast between northern and southern sourced deep waters was markedly greater than in the modern ocean. High density stratification could help explain the marked contrast in carbon isotope distribution recorded in the LGM ocean relative to that we observe today, but what made the ocean's density structure so different at the LGM? How did it evolve from one state to another? Further, given the sparsity of the LGM temperature and salinity data set, what else can we learn by increasing the spatial density of proxy records?

We investigate the cause and feasibility of a highly and salinity stratified deep ocean at the LGM and we work to increase the amount of information we can glean about the past ocean from pore fluid profiles of oxygen isotopes and chloride. Using a coupled ocean--sea ice--ice shelf cavity model we test whether the deep ocean density structure at the LGM can be explained by ice--ocean interactions over the Antarctic continental shelves, and show that a large contribution of the LGM salinity stratification can be explained through lower ocean temperature. In order to extract the maximum information from pore fluid profiles of oxygen isotopes and chloride we evaluate several inverse methods for ill-posed problems and their ability to recover bottom water histories from sediment pore fluid profiles. We demonstrate that Bayesian Markov Chain Monte Carlo parameter estimation techniques enable us to robustly recover the full solution space of bottom water histories, not only at the LGM, but through the most recent deglaciation and the Holocene up to the present. Finally, we evaluate a non-destructive pore fluid sampling technique, Rhizon samplers, in comparison to traditional squeezing methods and show that despite their promise, Rhizons are unlikely to be a good sampling tool for pore fluid measurements of oxygen isotopes and chloride.

Heuristics for Job-Shop Scheduling

Two methods of obtaining approximate solutions to the classic General Job-shop Scheduling Program are investigated. The first method is iterative. A sampling of the solution space is used to decide which of a collection of space pruning constraints are consistent with "good" schedules. The selected space pruning constraints are then used to reduce the search space and the sampling is repeated. This approach can be used either to verify whether some set of space pruning constraints can prune with discrimination or to generate solutions directly. Schedules can be represented as trajectories through a Cartesian space. Under the objective criteria of Minimum maximum Lateness family of "good" schedules (trajectories) are geometric neighbors (reside with some "tube") in this space. This second method of generating solutions takes advantage of this adjacency by pruning the space from the outside in thus converging gradually upon this "tube." One the average this methods significantly outperforms an array of the Priority Dispatch rules when the object criteria is that of Minimum Maximum Lateness. It also compares favorably with a recent relaxation procedure.

Articulated Pose Estimation in a Learned Smooth Space of Feasible Solutions

A learning based framework is proposed for estimating human body pose from a single image. Given a differentiable function that maps from pose space to image feature space, the goal is to invert the process: estimate the pose given only image features. The inversion is an ill-posed problem as the inverse mapping is a one to many process. Hence multiple solutions exist, and it is desirable to restrict the solution space to a smaller subset of feasible solutions. For example, not all human body poses are feasible due to anthropometric constraints. Since the space of feasible solutions may not admit a closed form description, the proposed framework seeks to exploit machine learning techniques to learn an approximation that is smoothly parameterized over such a space. One such technique is Gaussian Process Latent Variable Modelling. Scaled conjugate gradient is then used find the best matching pose in the space of feasible solutions when given an input image. The formulation allows easy incorporation of various constraints, e.g. temporal consistency and anthropometric constraints. The performance of the proposed approach is evaluated in the task of upper-body pose estimation from silhouettes and compared with the Specialized Mapping Architecture. The estimation accuracy of the Specialized Mapping Architecture is at least one standard deviation worse than the proposed approach in the experiments with synthetic data. In experiments with real video of humans performing gestures, the proposed approach produces qualitatively better estimation results.

An improved TLBO with elite strategy for parameters identification of PEM fuel cell and solar cell models

Clean and renewable energy generation and supply has drawn much attention worldwide in recent years, the proton exchange membrane (PEM) fuel cells and solar cells are among the most popular technologies. Accurately modeling the PEM fuel cells as well as solar cells is critical in their applications, and this involves the identification and optimization of model parameters. This is however challenging due to the highly nonlinear and complex nature of the models. In particular for PEM fuel cells, the model has to be optimized under different operation conditions, thus making the solution space extremely complex. In this paper, an improved and simplified teaching-learning based optimization algorithm (STLBO) is proposed to identify and optimize parameters for these two types of cell models. This is achieved by introducing an elite strategy to improve the quality of population and a local search is employed to further enhance the performance of the global best solution. To improve the diversity of the local search a chaotic map is also introduced. Compared with the basic TLBO, the structure of the proposed algorithm is much simplified and the searching ability is significantly enhanced. The performance of the proposed STLBO is firstly tested and verified on two low dimension decomposable problems and twelve large scale benchmark functions, then on the parameter identification of PEM fuel cell as well as solar cell models. Intensive experimental simulations show that the proposed STLBO exhibits excellent performance in terms of the accuracy and speed, in comparison with those reported in the literature.

Examination timetabling automation using hybrid meta-heuristics

Trabalho de projeto realizado para obtenção do grau de Mestre em Engenharia Informática e de Computadores

Combining extension matrix and integer programming for optimal concept learning

This paper proposes two integer programming models and their GA-based solutions for optimal concept learning. The models are built to obtain the optimal concept description in the form of propositional logic formulas from examples based on completeness, consistency and simplicity. The simplicity of the propositional rules is selected as the objective function of the integer programming models, and the completeness and consistency of the concept are used as the constraints. Considering the real-world problems that certain level of noise is contained in data set, the constraints in model 11 are slacked by adding slack-variables. To solve the integer programming models, genetic algorithm is employed to search the global solution space. We call our approach IP-AE. Its effectiveness is verified by comparing the experimental results with other well- known concept learning algorithms: AQ15 and C4.5.

Fuzzy ARTMAP and a hybrid chaos genetic algorithm for medical pattern classification

In this paper, an Evolutionary Artificial Neural Network (EANN), which combines the Fuzzy ARTMAP (FAM) neural network and a hybrid Chaos Genetic Algorithm (CGA), is proposed for undertaking pattern classification tasks. The hybrid CGA is a modified version of the hybrid real-coded genetic algorithms that includes a Chaotic Mapping Operator (CMO) in its search and adaptation process. It is used to evolve the connection weights in FAM, and the resulting EANN is known as FAM-hybrid CGA. The CMO in the hybrid CGA is used to generate a group of chromosomes that incorporates the characteristics of chaos. The chromosomes are then adapted with an arbitrary small amount of variation in every generation. As the evolution procedure proceeds, chromosomes with considerable differences are produced. Such chromosomes, which are located at different regions of interest in the solution space, are able to provide good solutions to undertake search and adaption problems. The effectiveness of the proposed FAM-hybrid CGA model is first evaluated using benchmark medical data sets from the UCI machine learning repository. Its applicability to medical decision support is then demonstrated using a real database of patient records with suspected Acute Coronary Syndrome. The results indicate that FAM-hybrid CGA is able to outperform its neural network counterpart (i.e., FAM), and it can be employed as a useful pattern classification tool for tackling medical decision support tasks.

Evolutionary optimization: a big data perspective

Stochastic search techniques such as evolutionary algorithms (EA) are known to be better explorer of search space as compared to conventional techniques including deterministic methods. However, in the era of big data like most other search methods and learning algorithms, suitability of evolutionary algorithms is naturally questioned. Big data pose new computational challenges including very high dimensionality and sparseness of data. Evolutionary algorithms' superior exploration skills should make them promising candidates for handling optimization problems involving big data. High dimensional problems introduce added complexity to the search space. However, EAs need to be enhanced to ensure that majority of the potential winner solutions gets the chance to survive and mature. In this paper we present an evolutionary algorithm with enhanced ability to deal with the problems of high dimensionality and sparseness of data. In addition to an informed exploration of the solution space, this technique balances exploration and exploitation using a hierarchical multi-population approach. The proposed model uses informed genetic operators to introduce diversity by expanding the scope of search process at the expense of redundant less promising members of the population. Next phase of the algorithm attempts to deal with the problem of high dimensionality by ensuring broader and more exhaustive search and preventing premature death of potential solutions. To achieve this, in addition to the above exploration controlling mechanism, a multi-tier hierarchical architecture is employed, where, in separate layers, the less fit isolated individuals evolve in dynamic sub-populations that coexist alongside the original or main population. Evaluation of the proposed technique on well known benchmark problems ascertains its superior performance. The algorithm has also been successfully applied to a real world problem of financial portfolio management. Although the proposed method cannot be considered big data-ready, it is certainly a move in the right direction.

Algoritmo Q-learning como estratégia de exploração e/ou explotação para metaheurísticas GRASP e algoritmo genético

Techniques of optimization known as metaheuristics have achieved success in the resolution of many problems classified as NP-Hard. These methods use non deterministic approaches that reach very good solutions which, however, don t guarantee the determination of the global optimum. Beyond the inherent difficulties related to the complexity that characterizes the optimization problems, the metaheuristics still face the dilemma of xploration/exploitation, which consists of choosing between a greedy search and a wider exploration of the solution space. A way to guide such algorithms during the searching of better solutions is supplying them with more knowledge of the problem through the use of a intelligent agent, able to recognize promising regions and also identify when they should diversify the direction of the search. This way, this work proposes the use of Reinforcement Learning technique - Q-learning Algorithm - as exploration/exploitation strategy for the metaheuristics GRASP (Greedy Randomized Adaptive Search Procedure) and Genetic Algorithm. The GRASP metaheuristic uses Q-learning instead of the traditional greedy-random algorithm in the construction phase. This replacement has the purpose of improving the quality of the initial solutions that are used in the local search phase of the GRASP, and also provides for the metaheuristic an adaptive memory mechanism that allows the reuse of good previous decisions and also avoids the repetition of bad decisions. In the Genetic Algorithm, the Q-learning algorithm was used to generate an initial population of high fitness, and after a determined number of generations, where the rate of diversity of the population is less than a certain limit L, it also was applied to supply one of the parents to be used in the genetic crossover operator. Another significant change in the hybrid genetic algorithm is the proposal of a mutually interactive cooperation process between the genetic operators and the Q-learning algorithm. In this interactive/cooperative process, the Q-learning algorithm receives an additional update in the matrix of Q-values based on the current best solution of the Genetic Algorithm. The computational experiments presented in this thesis compares the results obtained with the implementation of traditional versions of GRASP metaheuristic and Genetic Algorithm, with those obtained using the proposed hybrid methods. Both algorithms had been applied successfully to the symmetrical Traveling Salesman Problem, which was modeled as a Markov decision process

Arquitetura multiagente baseada em nuvem de partículas para hibridização de metaheurísticas

This thesis proposes an architecture of a new multiagent system framework for hybridization of metaheuristics inspired on the general Particle Swarm Optimization framework (PSO). The main contribution is to propose an effective approach to solve hard combinatory optimization problems. The choice of PSO as inspiration was given because it is inherently multiagent, allowing explore the features of multiagent systems, such as learning and cooperation techniques. In the proposed architecture, particles are autonomous agents with memory and methods for learning and making decisions, using search strategies to move in the solution space. The concepts of position and velocity originally defined in PSO are redefined for this approach. The proposed architecture was applied to the Traveling Salesman Problem and to the Quadratic Assignment Problem, and computational experiments were performed for testing its effectiveness. The experimental results were promising, with satisfactory performance, whereas the potential of the proposed architecture has not been fully explored. For further researches, the proposed approach will be also applied to multiobjective combinatorial optimization problems, which are closer to real-world problems. In the context of applied research, we intend to work with both students at the undergraduate level and a technical level in the implementation of the proposed architecture in real-world problems

Algoritmos científicos

The Scientific Algorithms are a new metaheuristics inspired in the scientific research process. The new method introduces the idea of theme to search the solution space of hard problems. The inspiration for this class of algorithms comes from the act of researching that comprises thinking, knowledge sharing and disclosing new ideas. The ideas of the new method are illustrated in the Traveling Salesman Problem. A computational experiment applies the proposed approach to a new variant of the Traveling Salesman Problem named Car Renter Salesman Problem. The results are compared to state-of-the-art algorithms for the latter problem

Stochastic scheduling - A flow shop no wait application and solutions

Minimizing the makespan of a flow-shop no-wait (FSNW) schedule where the processing times are randomly distributed is an important NP-Complete Combinatorial Optimization Problem. In spite of this, it can be found only in very few papers in the literature. By considering the Start Interval Concept, this problem can be formulated, in a practical way, in function of the probability of the success in preserve FSNW constraints for all tasks execution. With this formulation, for the particular case with 3 machines, this paper presents different heuristics solutions: by integrating local optimization steps with insertion procedures and by using genetic algorithms for search the solution space. Computational results and performance evaluations are commented. Copyright (C) 1998 IFAC.

Planejamento da expansão de sistemas de transmissão usando a metaheurística de busca em vizinhança variável

Pós-graduação em Engenharia Elétrica - FEIS