RESUMO

In this paper, a novel strategy to perform high-dimensional feature selection using an evolutionary algorithm for the automatic classification of coronary stenosis is introduced. The method involves a feature extraction stage to form a bank of 473 features considering different types such as intensity, texture and shape. The feature selection task is carried out on a high-dimensional feature bank, where the search space is denoted by O(2n) and n=473. The proposed evolutionary search strategy was compared in terms of the Jaccard coefficient and accuracy classification with different state-of-the-art methods. The highest feature selection rate, along with the best classification performance, was obtained with a subset of four features, representing a 99% discrimination rate. In the last stage, the feature subset was used as input to train a support vector machine using an independent testing set. The classification of coronary stenosis cases involves a binary classification type by considering positive and negative classes. The highest classification performance was obtained with the four-feature subset in terms of accuracy (0.86) and Jaccard coefficient (0.75) metrics. In addition, a second dataset containing 2788 instances was formed from a public image database, obtaining an accuracy of 0.89 and a Jaccard Coefficient of 0.80. Finally, based on the performance achieved with the four-feature subset, they can be suitable for use in a clinical decision support system.

RESUMO

The Product Line Architecture (PLA) is a crucial artifact for the development of Software Product Lines. However, PLA is a complex artifact to be designed due to its large size and the multiple conflicting properties that need to be considered to ensure its quality, requiring a great effort for the architect. PLA designing has been formulated as an optimization problem aiming at improving some architectural properties in order to maximize both the feature modularization and the relational cohesion, and to minimize the class coupling. This kind of problem was successfully solved by multi-objective evolutionary algorithm. Nevertheless, most of existing approaches optimize PLA designs without applying the crossover operator, one of the fundamental genetic operators. To overcome these limitations, this paper aims to intensify the search-based PLA design optimization by presenting three crossover operators. These operators were empirically evaluated in quantitative and qualitative studies using three well-studied PLA designs. The experiments were conducted with eight experimental configurations of NSGA-II in comparison with a baseline that uses only mutation operators. Empirical results showed that there are significant differences among the use of only mutation and mutation with crossover. Also, we observed that the crossover operators contributed to generate solutions with better feature modularization. Finally, we could see that the proposed operators complement each other, since the experiment that combines at least two of the proposed operators achieved better results.

RESUMO

Job Shop Scheduling is currently one of the most addressed planning and scheduling optimization problems in the field. Due to its complexity, as it belongs to the NP-Hard class of problems, meta-heuristics are one of the most commonly used approaches in its resolution, with Genetic Algorithms being one of the most effective methods in this category. However, it is well known that this meta-heuristic is affected by phenomena that worsen the quality of its population, such as premature convergence and population concentration in regions of local optima. To circumvent these difficulties, we propose, in this work, the use of a guidance operator responsible for modifying ill-adapted individuals using genetic material from well-adapted individuals. We also propose, in this paper, a new method of determining the genetic quality of individuals using genetic frequency analysis. Our method is evaluated over a wide range of modern GAs and considers two case studies defined by well-established JSSP benchmarks in the literature. The results show that the use of the proposed operator assists in managing individuals with poor fitness values, which improves the population quality of the algorithms and, consequently, leads to obtaining better results in the solution of JSSP instances. Finally, the use of the proposed operator in the most elaborate GA-like method in the literature was able to reduce its mean relative error from 1.395% to 0.755%, representing an improvement of 45.88%.

Assuntos

RESUMO

A critical factor in the logistic management of firms is the degree of efficiency of the operations in distribution centers. Of particular interest is the pick-up process, since it is the costliest operation, amounting to 50 and up to 75% of the total cost of the activities in storage facilities. In this paper we jointly address the order batching problem (OBP) and the order picking problem (OPP). The former problem amounts to find optimal batches of goods to be picked up, by restructuring incoming orders by either splitting up large orders or combining small orders into larger ones that can then be picked in a single picking tour. The OPP, in turn, involves identifying optimal sequences of visits to the storage positions in which the goods to be included in each batch are stored. We seek to design a plan that minimizes the total operational cost of the pick-up process, proportional to the displacement times around the storage area as well as to all the time spent in pick-ups and finishing up orders to be punctually delivered. Earliness or tardiness will induce inefficiency costs, be it because of the excessive use of space or breaches of contracts with customers. Tsai, Liou and Huang in 2008 have generated 2D and 3D instances. In previous works we have addressed the 2D ones, achieving very good results. Here we focus on 3D instances (the articles are placed at different levels in the storage center), which involve a higher complexity. This contributes to improve the performance of the hybrid evolutionary algorithm (HEA) applied in our previous works.

Assuntos

RESUMO

Computational peptide design is useful for therapeutics, diagnostics, and vaccine development. To select the most promising peptide candidates, the key is describing accurately the peptide-target interactions at the molecular level. We here review a computational peptide design protocol whose key feature is the use of all-atom explicit solvent molecular dynamics for describing the different peptide-target complexes explored during the optimization. We describe the milestones behind the development of this protocol, which is now implemented in an open-source code called PARCE. We provide a basic tutorial to run the code for an antibody fragment design example. Finally, we describe three additional applications of the method to design peptides for different targets, illustrating the broad scope of the proposed approach.

Assuntos

RESUMO

The planning of hospital beds is among the most debated problems in healthcare. Despite being an important issue, many initiatives have failed to sustain services improvements, resulting in high costs and also high refusal rates. The stochastic problem involves conflicting criteria, therefore, we propose a Simulation-Optimisation approach to solve it. The Evolutionary Algorithm NSGA-II drives the process, and the solutions are validated and evaluated via Discrete Event Simulation. An application is performed in one of the health regions of the state of Minas Gerais, Brazil, where the public health system assists nearly 80% of the patients. The results pointed out that the proposed approach could find efficient and feasible solutions for the problem. Therefore, it is a good alternative to empirical methods currently used in Brazil to set hospital beds allocation.

Assuntos

RESUMO

AIMS: Simple surrogate indices of insulin sensitivity have been conceived to deal with costly and complicated approaches, such as the hyperinsulinemic-euglycemic clamp; however, their use has not been widespread given their variabilities in different populations. In this paper, we present two simple surrogate indices, one that uses fasting glucose and insulin values and the other based on the values from the oral glucose tolerance test. MATERIALS AND METHODS: The proposed methods integrate easy-to-obtain anthropometric measures. Evolutionary algorithms were used to optimize the proposed methods by maximizing its correlation with the Stumvoll MCR method. RESULTS AND CONCLUSION: When the proposed indices were applied to three study groups (control subjects, metabolic syndrome, marathon runners), a reduction in the intergroup variability of the insulin sensitivity was obtained. Moreover, the proposed index based on the oral glucose tolerance test (OGTT), which considers the glucose metabolism process and the hepatic and peripheral insulin sensitivity, showed stronger correlations with the Stumvoll method and lower intergroup variability than the fasting one.

Assuntos

RESUMO

BACKGROUND: Antimicrobial peptides are a promising alternative for combating pathogens resistant to conventional antibiotics. Computer-assisted peptide discovery strategies are necessary to automatically assess a significant amount of data by generating models that efficiently classify what an antimicrobial peptide is, before its evaluation in the wet lab. Model's performance depends on the selection of molecular descriptors for which an efficient and effective approach has recently been proposed. Unfortunately, how to adapt this method to the selection of molecular descriptors for the classification of antimicrobial peptides and the performance it can achieve, have only preliminary been explored. RESULTS: We propose an adaptation of this successful feature selection approach for the weighting of molecular descriptors and assess its performance. The evaluation is conducted on six high-quality benchmark datasets that have previously been used for the empirical evaluation of state-of-art antimicrobial prediction tools in an unbiased manner. The results indicate that our approach substantially reduces the number of required molecular descriptors, improving, at the same time, the performance of classification with respect to using all molecular descriptors. Our models also outperform state-of-art prediction tools for the classification of antimicrobial and antibacterial peptides. CONCLUSIONS: The proposed methodology is an efficient approach for the development of models to classify antimicrobial peptides. Particularly in the generation of models for discrimination against a specific antimicrobial activity, such as antibacterial. One of our future directions is aimed at using the obtained classifier to search for antimicrobial peptides in various transcriptomes.

Assuntos