964 resultados para Labeling
Resumo:
Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
Resumo:
Both Semi-Supervised Leaning and Active Learning are techniques used when unlabeled data is abundant, but the process of labeling them is expensive and/or time consuming. In this paper, those two machine learning techniques are combined into a single nature-inspired method. It features particles walking on a network built from the data set, using a unique random-greedy rule to select neighbors to visit. The particles, which have both competitive and cooperative behavior, are created on the network as the result of label queries. They may be created as the algorithm executes and only nodes affected by the new particles have to be updated. Therefore, it saves execution time compared to traditional active learning frameworks, in which the learning algorithm has to be executed several times. The data items to be queried are select based on information extracted from the nodes and particles temporal dynamics. Two different rules for queries are explored in this paper, one of them is based on querying by uncertainty approaches and the other is based on data and labeled nodes distribution. Each of them may perform better than the other according to some data sets peculiarities. Experimental results on some real-world data sets are provided, and the proposed method outperforms the semi-supervised learning method, from which it is derived, in all of them.
Resumo:
Concept drift, which refers to non stationary learning problems over time, has increasing importance in machine learning and data mining. Many concept drift applications require fast response, which means an algorithm must always be (re)trained with the latest available data. But the process of data labeling is usually expensive and/or time consuming when compared to acquisition of unlabeled data, thus usually only a small fraction of the incoming data may be effectively labeled. Semi-supervised learning methods may help in this scenario, as they use both labeled and unlabeled data in the training process. However, most of them are based on assumptions that the data is static. Therefore, semi-supervised learning with concept drifts is still an open challenging task in machine learning. Recently, a particle competition and cooperation approach has been developed to realize graph-based semi-supervised learning from static data. We have extend that approach to handle data streams and concept drift. The result is a passive algorithm which uses a single classifier approach, naturally adapted to concept changes without any explicit drift detection mechanism. It has built-in mechanisms that provide a natural way of learning from new data, gradually "forgetting" older knowledge as older data items are no longer useful for the classification of newer data items. The proposed algorithm is applied to the KDD Cup 1999 Data of network intrusion, showing its effectiveness.
Resumo:
Endo-oligopeptidase A, EC 3.4.22.19, converts small enkephalin-containing peptides into the corresponding enkephalins in vitro. We investigated the presence of endooligopeptidase A in the retina and its possible colocalization with enkephalins in retinal neurons. The specific activity of endo-oligopeptidase A found in pigeon retinae (30.3 +/- 7.3 mU/mg, mean +/- standard deviation) was four times higher than in rabbit retinae (7.0 +/- 1.1 mU/mg). The enzyme activity was not modified by EDTA, but it was enhanced by dithiothreitol and inhibited by zinc and 5,5'-dithiobis(2-nitrobenzoic acid). Immunohistochemical experiments with a purified antiserum against rabbit endo-oligopeptidase A revealed labeled neurons in both the inner nuclear layer and the ganglion cell layer of pigeon and rabbit retinae. Double-labeling immunofluorescence experiments demonstrated that about 90% of neurons containing endo-oligopeptidase A-like immunoreactivity also contained [Leu5]-enkephalin-like immunoreactivity. These colocalization results may represent an important step toward the demonstration of the possible involvement of endo-oligopeptidase A in enkephalin generation in vivo.
Resumo:
Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
Resumo:
The discussions about the energy rationalization are gaining more and more space on the everyday of engineers and other professionals in the field of energy. A greater focus is being tied to commercial buildings, because they are one of the biggest responsible for the high energetic expenditure according to the National Energetic Balance, especially by the use of air conditioners for the people’s thermal comfort. Such finding made INMETRO to develop a building labeling procedure through th National Program of Electric Energy Conservation in Buildings; by means of this systematic, the built areas are classified by levels that go from A(the best) to E(the worst), taking into account envelopment, illumination system and air conditioning. A bonus process based on water rationalization, use of renewable energy sources, cogeneration systems or technical innovations, allows the classification to be changed up to one level. A study made exactly on a commercial building, the building of Foundation for Technological and Scientific Development located inside of the College of Engineering of Guaratinguetá, sought to identify technical/economic alternatives for the building certification and appealed to the bonus based on water rationalization. The present study is based on analysis of deploying a cogeneration system formed by internal combustion engine and an absorption refrigeration system as bonus alternative, so that the costs are analyzed regarding energy consumption and use of the motor. With the calculations and results, the viability of the building to receive a bonus from INMETRO through its National Program for Energy Conservation in Buildings is studied and concluded from this study whether or not you can get a better level of energy efficiency for the building in question... (Complete abstract click electronic access below)
Resumo:
This paper presents a study on the labeling Procel Build, an action plan for energy efficiency in buildings which aims to build the foundations necessary to rationalize energy consumption in buildings in Brazil (PROCEL, 2009a, p. 5), ie Procel builds aims to ensure the construction of the building that has a low energy expenditure, but still provide a comfortable environment. In this research, it is also shown how the labels are obtained, which may be general or partial. The reader will find further explanation of the three aspects of a building, and they Envelopment, Lighting system and Air Conditioning System Finally, we present the advantages and challenges of system builds procel, showing the need for greater investment and disclosure of elements that contribute to the implementation of sustainable buildings in Brazil
Resumo:
The food industry has been rapidly modernized, and with this the disposal and the consumption of industrialized food has been increasing continually. These products lose their original morphological characteristics, requiring fast and reliable tests that could help to identify the species in question, as most fraudulent behavior in the milk and dairy industry (meat and fish) is carried out where there is partial or total exchange of the original material for other with less value at market. Nowadays there is a lot of techniques that can be used for the identification of animal species, based on muscle protein, or DNA analysis. In the case of protein based analysis, we can mention several types of electrophoresis and immunologic methods, as ELISA. In the case of DNA based methods, we have several assays that use the amplification of DNA fragments, known as PCR, as proof. All these techniques have advantages and disadvantages that can be affected by factors- the sample condition, or the degree of relation between the species in question. Because of this, it’s necessary that a continuous study looking for the improvement of the available techniques, making sure that the confirmation of food authenticity is in place. This is to ensure the true product value, to comply with labeling regulationand and protect the consumer of frauds
Resumo:
The study aims at analyzing the potential impacts of environmental labeling programs, emphasizing the European Community program on Brazilian exports to the EU. For this, we traced the historical growth of bilateral relations between the EU and Brazil, which shows a strategic partnership, and then it was listed trade barriers (tariff and nontariff) imposed by these two markets to the international trade relations. Later, the emergence of environmental seals was discussed, as well as their potential barriers to trade. Finally, selected products were analyzed in order to know the potential effects of environmental labeling on products exported by Brazil
Resumo:
Pós-graduação em Cirurgia Veterinária - FCAV
Resumo:
Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
Resumo:
Endometriosis is a continuous and progressive disease with a poorly understood aetiology, pathophysiology and natural history. This study evaluated the histological differences between eutopic and ectopic endometria (abdominal wall endometriosis) and the expression of mast cell proteases (tryptase and chymase), annexin A1 (ANXA1) and formyl peptide receptor 1 (FPR1). Ectopic endometrium from 18 women with abdominal wall endometriosis and eutopic endometrium from 10 women without endometriosis were obtained. The endometrial samples were analysed by histopathology, immunohistochemistry and ultrastructural immunogold labeling to determine mast cell heterogeneity (tryptase and chymase positive cells) and the expression levels of ANXA1 and FPR1. Histopathological analysis of the endometriotic lesions showed a glandular pattern of mixed differentiation and an undifferentiated morphology with a significant influx of inflammatory cells and a change in mast cell heterogeneity, as evidenced by a significant increase in the number of chymase-positive cells and endogenous chymase expression. The undifferentiated glandular pattern of endometriotic lesions was positively associated with a marked increase and co-localization of ANXA1 and FPR1 in the epithelial cells. In conclusion, the co-upregulated expression of mast cell chymase and ANXA1–FPR1 system in ectopic endometrium suggests their involvement in the development of endometriotic lesions.
Resumo:
Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
Resumo:
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
Resumo:
Pós-graduação em Ciência e Tecnologia Animal - FEIS