Particle competition and cooperation to prevent error propagation from mislabeled data in semi-supervised learning

Semi-supervised learning is applied to classification problems where only a small portion of the data items is labeled. In these cases, the reliability of the labels is a crucial factor, because mislabeled items may propagate wrong labels to a large portion or even the entire data set. This paper aims to address this problem by presenting a graph-based (network-based) semi-supervised learning method, specifically designed to handle data sets with mislabeled samples. The method uses teams of walking particles, with competitive and cooperative behavior, for label propagation in the network constructed from the input data set. The proposed model is nature-inspired and it incorporates some features to make it robust to a considerable amount of mislabeled data items. Computer simulations show the performance of the method in the presence of different percentage of mislabeled data, in networks of different sizes and average node degree. Importantly, these simulations reveals the existence of the critical points of the mislabeled subset size, below which the network is free of wrong label contamination, but above which the mislabeled samples start to propagate their labels to the rest of the network. Moreover, numerical comparisons have been made among the proposed method and other representative graph-based semi-supervised learning methods using both artificial and real-world data sets. Interestingly, the proposed method has increasing better performance than the others as the percentage of mislabeled samples is getting larger. © 2012 IEEE.

Risk management in software projects through Knowledge Management techniques: Cases in Brazilian Incubated Technology-Based Firms

In businesses such as the software industry, which uses knowledge as a resource, activities are knowledge intensive, requiring constant adoption of new technologies and practices. Another feature of this environment is that the industry is particularly susceptible to failure; with this in mind, the objective of this research is to analyze the integration of Knowledge Management techniques into the activity of risk management as it applies to software development projects of micro and small Brazilian incubated technology-based firms. Research methods chosen were the Multiple Case Study. The main risk factor for managers and developers is that scope or goals are often unclear or misinterpreted. For risk management, firms have found that Knowledge Management techniques of conversion combination would be the most applicable for use; however, those most commonly used refer to the conversion mode as internalization.. © 2013 Elsevier Ltd. APM and IPMA.

Structural and Phylogenetic Studies with MjTX-I Reveal a Multi-Oligomeric Toxin - a Novel Feature in Lys49-PLA2s Protein Class

The mortality caused by snakebites is more damaging than many tropical diseases, such as dengue haemorrhagic fever, cholera, leishmaniasis, schistosomiasis and Chagas disease. For this reason, snakebite envenoming adversely affects health services of tropical and subtropical countries and is recognized as a neglected disease by the World Health Organization. One of the main components of snake venoms is the Lys49-phospholipases A2, which is catalytically inactive but possesses other toxic and pharmacological activities. Preliminary studies with MjTX-I from Bothrops moojeni snake venom revealed intriguing new structural and functional characteristics compared to other bothropic Lys49-PLA2s. We present in this article a comprehensive study with MjTX-I using several techniques, including crystallography, small angle X-ray scattering, analytical size-exclusion chromatography, dynamic light scattering, myographic studies, bioinformatics and molecular phylogenetic analyses.Based in all these experiments we demonstrated that MjTX-I is probably a unique Lys49-PLA2, which may adopt different oligomeric forms depending on the physical-chemical environment. Furthermore, we showed that its myotoxic activity is dramatically low compared to other Lys49-PLA2s, probably due to the novel oligomeric conformations and important mutations in the C-terminal region of the protein. The phylogenetic analysis also showed that this toxin is clearly distinct from other bothropic Lys49-PLA2s, in conformity with the peculiar oligomeric characteristics of MjTX-I and possible emergence of new functionalities inresponse to environmental changes and adaptation to new preys. © 2013 Salvador et al.

Exterior orientation of CBERS-2B imagery using multi-feature control and orbital data

The major contribution of this paper relates to the practical advantages of combining Ground Control Points (GCPs), Ground Control Lines (GCLs) and orbital data to estimate the exterior orientation parameters of images collected by CBERS-2B (China-Brazil Earth Resources Satellite) HRC (High-resolution Camera) and CCD (High-resolution CCD Camera) sensors. Although the CBERS-2B is no longer operational, its images are still being used in Brazil, and the next generations of the CBERS satellite will have sensors with similar technical features, which motivates the study presented in this paper. The mathematical models that relate the object and image spaces are based on collinearity (for points) and coplanarity (for lines) conditions. These models were created in an in-house developed software package called TMS (Triangulation with Multiple Sensors) with multi-feature control (GCPs and GCLs). Experiments on a block of four CBERS-2B HRC images and on one CBERS-2B CCD image were performed using both models. It was observed that the combination of GCPs and GCLs provided better bundle block adjustment results than conventional bundle adjustment using only GCPs. The results also demonstrate the advantages of using primarily orbital data when the number of control entities is reduced. © 2013 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS).

ECG arrhythmia classification based on optimum-path forest

An important tool for the heart disease diagnosis is the analysis of electrocardiogram (ECG) signals, since the non-invasive nature and simplicity of the ECG exam. According to the application, ECG data analysis consists of steps such as preprocessing, segmentation, feature extraction and classification aiming to detect cardiac arrhythmias (i.e.; cardiac rhythm abnormalities). Aiming to made a fast and accurate cardiac arrhythmia signal classification process, we apply and analyze a recent and robust supervised graph-based pattern recognition technique, the optimum-path forest (OPF) classifier. To the best of our knowledge, it is the first time that OPF classifier is used to the ECG heartbeat signal classification task. We then compare the performance (in terms of training and testing time, accuracy, specificity, and sensitivity) of the OPF classifier to the ones of other three well-known expert system classifiers, i.e.; support vector machine (SVM), Bayesian and multilayer artificial neural network (MLP), using features extracted from six main approaches considered in literature for ECG arrhythmia analysis. In our experiments, we use the MIT-BIH Arrhythmia Database and the evaluation protocol recommended by The Association for the Advancement of Medical Instrumentation. A discussion on the obtained results shows that OPF classifier presents a robust performance, i.e.; there is no need for parameter setup, as well as a high accuracy at an extremely low computational cost. Moreover, in average, the OPF classifier yielded greater performance than the MLP and SVM classifiers in terms of classification time and accuracy, and to produce quite similar performance to the Bayesian classifier, showing to be a promising technique for ECG signal analysis. © 2012 Elsevier Ltd. All rights reserved.

Irregular satellites of Jupiter: Three-dimensional study of binary-asteroid captures

Among the hidden pieces of the giant puzzle, which is our Solar system, the origins of irregularsatellites of the giant planets stand to be explained, while the origins of regular satellites arewell explained by the in situ formation model through matter accretion. Once they are notlocally formed, the most acceptable theory predicts that they had been formed elsewhere andbecame captured later, most likely during the last stage of planet formation. However, underthe restricted three-body problem theory, captures are temporary and there is still no assistedcapture mechanism which is well established. In a previous work, we showed that the capturemechanism of a binary asteroid under the co-planar four-body scenario yielded permanentcaptured objects with an orbital shape which is very similar to those of the actual progradeirregular Jovian satellites. By extending our previous study to a 3D case, here we demonstratethat the capture mechanism of a binary asteroid can produce permanent captures of objects byitself which have very similar orbits to irregular Jovian satellites. Some of the captured objectswithout aid of gas drag or other mechanisms present a triplet: semi-major axis, eccentricityand inclination, which is comparable to the already known irregular Jovian objects. © 2013 The Authors Published by Oxford University Press on behalf of the Royal Astronomical Society.

Unsupervised segmentation method for cuboidal cell nuclei in histological prostate images based on minimum cross entropy

This paper presents a novel segmentation method for cuboidal cell nuclei in images of prostate tissue stained with hematoxylin and eosin. The proposed method allows segmenting normal, hyperplastic and cancerous prostate images in three steps: pre-processing, segmentation of cuboidal cell nuclei and post-processing. The pre-processing step consists of applying contrast stretching to the red (R) channel to highlight the contrast of cuboidal cell nuclei. The aim of the second step is to apply global thresholding based on minimum cross entropy to generate a binary image with candidate regions for cuboidal cell nuclei. In the post-processing step, false positives are removed using the connected component method. The proposed segmentation method was applied to an image bank with 105 samples and measures of sensitivity, specificity and accuracy were compared with those provided by other segmentation approaches available in the specialized literature. The results are promising and demonstrate that the proposed method allows the segmentation of cuboidal cell nuclei with a mean accuracy of 97%. © 2013 Elsevier Ltd. All rights reserved.

Orientação semi-automática de imagens CBERS usando rodovias como controle de campo

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Avaliação da atividade tripanocida do ácido gálico e seus ésteres contra formas epimastigotas de Trypanosoma cruzi: Rogério Andréo. -

Pós-graduação em Biotecnologia - IQ

Metodologia para avaliação de rodovias extraídas computacionalmente em imagens digitais

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Extração semi-automática de edificações com análise do modelo numérico de elevações

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Entomofauna associada a goiabeira Psidium guajava L. em pomares experimentais comerciais em Vista Alegre do Alto – SP e semi-orgânicos em Pindorama – SP

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Tradução juramentada de contratos de compra e venda

Pós-graduação em Estudos Linguísticos - IBILCE

Helmintos parasitos de mocós (Kerodon rupestris Wied, 1820), de vida livre e de cativeiro, criados no semi-árido nordestino

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Capoeira Angola: saberes, valores e atitudes na formação do mestre

Pós-graduação em Ciências da Motricidade - IBRC