Precision-recall operating characteristic (P-ROC) curves in imprecise environments

Traditionally, machine learning algorithms have been evaluated in applications where assumptions can be reliably made about class priors and/or misclassification costs. In this paper, we consider the case of imprecise environments, where little may be known about these factors and they may well vary significantly when the system is applied. Specifically, the use of precision-recall analysis is investigated and compared to the more well known performance measures such as error-rate and the receiver operating characteristic (ROC). We argue that while ROC analysis is invariant to variations in class priors, this invariance in fact hides an important factor of the evaluation in imprecise environments. Therefore, we develop a generalised precision-recall analysis methodology in which variation due to prior class probabilities is incorporated into a multi-way analysis of variance (ANOVA). The increased sensitivity and reliability of this approach is demonstrated in a remote sensing application.

Hybrid Background Subtraction in Video using Bi-level CodeBook Model

Detection of Objects in Video is a highly demanding area of research. The Background Subtraction Algorithms can yield better results in Foreground Object Detection. This work presents a Hybrid CodeBook based Background Subtraction to extract the foreground ROI from the background. Codebooks are used to store compressed information by demanding lesser memory usage and high speedy processing. This Hybrid method which uses Block-Based and Pixel-Based Codebooks provide efficient detection results; the high speed processing capability of block based background subtraction as well as high Precision Rate of pixel based background subtraction are exploited to yield an efficient Background Subtraction System. The Block stage produces a coarse foreground area, which is then refined by the Pixel stage. The system’s performance is evaluated with different block sizes and with different block descriptors like 2D-DCT, FFT etc. The Experimental analysis based on statistical measurements yields precision, recall, similarity and F measure of the hybrid system as 88.74%, 91.09%, 81.66% and 89.90% respectively, and thus proves the efficiency of the novel system.

Performance Study of an Improved Legendre Moment Descriptor as Region-Based Shape Descriptor

This paper reports a novel region-based shape descriptor based on orthogonal Legendre moments. The preprocessing steps for invariance improvement of the proposed Improved Legendre Moment Descriptor (ILMD) are discussed. The performance of the ILMD is compared to the MPEG-7 approved region shape descriptor, angular radial transformation descriptor (ARTD), and the widely used Zernike moment descriptor (ZMD). Set B of the MPEG-7 CE-1 contour database and all the datasets of the MPEG-7 CE-2 region database were used for experimental validation. The average normalized modified retrieval rate (ANMRR) and precision- recall pair were employed for benchmarking the performance of the candidate descriptors. The ILMD has lower ANMRR values than ARTD for most of the datasets, and ARTD has a lower value compared to ZMD. This indicates that overall performance of the ILMD is better than that of ARTD and ZMD. This result is confirmed by the precision-recall test where ILMD was found to have better precision rates for most of the datasets tested. Besides retrieval accuracy, ILMD is more compact than ARTD and ZMD. The descriptor proposed is useful as a generic shape descriptor for content-based image retrieval (CBIR) applications

HTS and HTSn: New shape descriptors based on Hough transform statistics

With the widespread proliferation of computers, many human activities entail the use of automatic image analysis. The basic features used for image analysis include color, texture, and shape. In this paper, we propose a new shape description method, called Hough Transform Statistics (HTS), which uses statistics from the Hough space to characterize the shape of objects or regions in digital images. A modified version of this method, called Hough Transform Statistics neighborhood (HTSn), is also presented. Experiments carried out on three popular public image databases showed that the HTS and HTSn descriptors are robust, since they presented precision-recall results much better than several other well-known shape description methods. When compared to Beam Angle Statistics (BAS) method, a shape description method that inspired their development, both the HTS and the HTSn methods presented inferior results regarding the precision-recall criterion, but superior results in the processing time and multiscale separability criteria. The linear complexity of the HTS and the HTSn algorithms, in contrast to BAS, make them more appropriate for shape analysis in high-resolution image retrieval tasks when very large databases are used, which are very common nowadays. (C) 2014 Elsevier Inc. All rights reserved.

PROBABILISTIC MODELING OF RAINFALL INDUCED LANDSLIDE HAZARD ASSESSMENT IN SAN JUAN LA LAGUNA, SOLOLÁ, GUATEMALA

The municipality of San Juan La Laguna, Guatemala is home to approximately 5,200 people and located on the western side of the Lake Atitlán caldera. Steep slopes surround all but the eastern side of San Juan. The Lake Atitlán watershed is susceptible to many natural hazards, but most predictable are the landslides that can occur annually with each rainy season, especially during high-intensity events. Hurricane Stan hit Guatemala in October 2005; the resulting flooding and landslides devastated the Atitlán region. Locations of landslide and non-landslide points were obtained from field observations and orthophotos taken following Hurricane Stan. This study used data from multiple attributes, at every landslide and non-landslide point, and applied different multivariate analyses to optimize a model for landslides prediction during high-intensity precipitation events like Hurricane Stan. The attributes considered in this study are: geology, geomorphology, distance to faults and streams, land use, slope, aspect, curvature, plan curvature, profile curvature and topographic wetness index. The attributes were pre-evaluated for their ability to predict landslides using four different attribute evaluators, all available in the open source data mining software Weka: filtered subset, information gain, gain ratio and chi-squared. Three multivariate algorithms (decision tree J48, logistic regression and BayesNet) were optimized for landslide prediction using different attributes. The following statistical parameters were used to evaluate model accuracy: precision, recall, F measure and area under the receiver operating characteristic (ROC) curve. The algorithm BayesNet yielded the most accurate model and was used to build a probability map of landslide initiation points. The probability map developed in this study was also compared to the results of a bivariate landslide susceptibility analysis conducted for the watershed, encompassing Lake Atitlán and San Juan. Landslides from Tropical Storm Agatha 2010 were used to independently validate this study’s multivariate model and the bivariate model. The ultimate aim of this study is to share the methodology and results with municipal contacts from the author's time as a U.S. Peace Corps volunteer, to facilitate more effective future landslide hazard planning and mitigation.

On the practicality of global flow analysis of logic programs

This paper addresses the issue of the practicality of global flow analysis in logic program compilation, in terms of both speed and precision of analysis. It discusses design and implementation aspects of two practical abstract interpretation-based flow analysis systems: MA3, the MOO Andparallel Analyzer and Annotator; and Ms, an experimental mode inference system developed for SB-Prolog. The paper also provides performance data obtained from these implementations. Based on these results, it is concluded that the overhead of global flow analysis is not prohibitive, while the results of analysis can be quite precise and useful.

Colonic movements in healthy subjects as monitored by a Magnet Tracking System.

The Magnet Tracking System (MTS) is a minimally-invasive technique of continuous evaluation of gastrointestinal motility. In this study, MTS was used to analyse colonic propulsive dynamics and compare the transit of a magnetic pill with that of standard radio-opaque markers. MTS monitors the progress in real time of a magnetic pill through the gut. Ten men and 10 women with regular daily bowel movements swallowed this pill and 10 radio-opaque markers at 8 pm. Five hours of recordings were conducted during 2 following mornings. Origin, direction, amplitude and velocity of movements were analysed relative to space-time plots of the pill trajectory. Abdominal radiographs were taken to compare the progress of both pill and markers. The magnetic pill lay idle for 90% of its sojourn in the colon; its total retrograde displacement accounted for only 20% of its overall movement. Analysis of these movements showed a bimodal distribution of velocities: around 1.5 and 50 cm min(-1), the latter being responsible for 2/3 of distance traversed. There were more movements overall and more mass movements in males. Net hourly forward progress was greater in the left than right colon, and greater in males. The position of the magnetic pill correlated well with the advancement of markers. MTS showed patterns and propulsion dynamics of colonic segments with as yet unmet precision. Detailed analysis of slow and fast patterns of colonic progress makes it possible to specify the motility of colonic segments, and any variability in gender. Such analysis opens up promising avenues in studies of motility disorders.

Development and standardization of an indirect ELISA for the serological diagnosis of classical swine fever

An indirect enzyme linked immunoassay (ELISA-I) was developed and standardized for the serological diagnosis of classical swine fever (CSF). For the comparison, nine hundred and thirty-seven swine serum samples were tested by serum neutralization followed by immunoperoxidase staining (NPLA), considered as the standard. Of these, 223 were positive and 714 negative for neutralizing antibodies to classical swine fever virus (CSFV). In relation to the NPLA, the ELISA-I presented a 98.2% sensitivity; 92.86% specificity, 81.11% positive predictive value, 99.4% negative predictive value and a 94.1% precision. Statistical analysis showed a very strong correlation (r=0,94) between both tests. When compared to a commercially available ELISA kit, the performance of both, in relation to the NPLA, was similar. It was concluded that the ELISA-I is suitable for large scale screening of antibodies to classical swine fever virus, although it does not distinguish antibodies to classical swine fever virus from those induced by other pestiviruses.

Equação de regressão linear múltipla para estimativa do erro experimental

Este trabalho teve por objetivo estimar equações de regressão linear múltipla tendo, como variáveis explicativas, as demais características avaliadas em experimento de milho e, como variáveis principais, a diferença mínima significativa em percentagem da média (DMS%) e quadrado médio do erro (QMe), para peso de grãos. Com 610 experimentos conduzidos na Rede de Ensaios Nacionais de Competição de Cultivares de Milho, realizados entre 1986 e 1996 (522 experimentos) e em 1997 (88 experimentos), estimaram-se duas equações de regressão, com os 522 experimentos, validando estas pela análise de regressão simples entre os valores reais e os estimados pelas equações, com os 88 restantes, observando que, para a DMS% a equação não estimava o mesmo valor que a fórmula original e, para o QMe, a equação poderia ser utilizada na estimação. Com o teste de Lilliefors, verificou-se que os valores do QMe aderiam à distribuição normal padrão e foi construída uma tabela de classificação dos valores do QMe, baseada nos valores observados na análise da variância dos experimentos e nos estimados pela equação de regressão.

An automatic method for segmentation of fission tracks in epidote crystal photomicrographs

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

Estimating the concentration of gold nanoparticles incorporated on natural rubber membranes using multi-level starlet optimal segmentation

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

Processamento de fotomicrografias por meio da transformada wavelet starlet

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

Exchange of nitrogen dioxide (NO 2) between plants and the atmosphere under laboratory and field conditions

Nitrogen is an essential nutrient. It is for human, animal and plants a constituent element of proteins and nucleic acids. Although the majority of the Earth’s atmosphere consists of elemental nitrogen (N2, 78 %) only a few microorganisms can use it directly. To be useful for higher plants and animals elemental nitrogen must be converted to a reactive oxidized form. This conversion happens within the nitrogen cycle by free-living microorganisms, symbiotic living Rhizobium bacteria or by lightning. Humans are able to synthesize reactive nitrogen through the Haber-Bosch process since the beginning of the 20th century. As a result food security of the world population could be improved noticeably. On the other side the increased nitrogen input results in acidification and eutrophication of ecosystems and in loss of biodiversity. Negative health effects arose for humans such as fine particulate matter and summer smog. Furthermore, reactive nitrogen plays a decisive role at atmospheric chemistry and global cycles of pollutants and nutritive substances.rnNitrogen monoxide (NO) and nitrogen dioxide (NO2) belong to the reactive trace gases and are grouped under the generic term NOx. They are important components of atmospheric oxidative processes and influence the lifetime of various less reactive greenhouse gases. NO and NO2 are generated amongst others at combustion process by oxidation of atmospheric nitrogen as well as by biological processes within soil. In atmosphere NO is converted very quickly into NO2. NO2 is than oxidized to nitrate (NO3-) and to nitric acid (HNO3), which bounds to aerosol particles. The bounded nitrate is finally washed out from atmosphere by dry and wet deposition. Catalytic reactions of NOx are an important part of atmospheric chemistry forming or decomposing tropospheric ozone (O3). In atmosphere NO, NO2 and O3 are in photosta¬tionary equilibrium, therefore it is referred as NO-NO2-O3 triad. At regions with elevated NO concentrations reactions with air pollutions can form NO2, altering equilibrium of ozone formation.rnThe essential nutrient nitrogen is taken up by plants mainly by dissolved NO3- entering the roots. Atmospheric nitrogen is oxidized to NO3- within soil via bacteria by nitrogen fixation or ammonium formation and nitrification. Additionally atmospheric NO2 uptake occurs directly by stomata. Inside the apoplast NO2 is disproportionated to nitrate and nitrite (NO2-), which can enter the plant metabolic processes. The enzymes nitrate and nitrite reductase convert nitrate and nitrite to ammonium (NH4+). NO2 gas exchange is controlled by pressure gradients inside the leaves, the stomatal aperture and leaf resistances. Plant stomatal regulation is affected by climate factors like light intensity, temperature and water vapor pressure deficit. rnThis thesis wants to contribute to the comprehension of the effects of vegetation in the atmospheric NO2 cycle and to discuss the NO2 compensation point concentration (mcomp,NO2). Therefore, NO2 exchange between the atmosphere and spruce (Picea abies) on leaf level was detected by a dynamic plant chamber system under labo¬ratory and field conditions. Measurements took place during the EGER project (June-July 2008). Additionally NO2 data collected during the ECHO project (July 2003) on oak (Quercus robur) were analyzed. The used measuring system allowed simultaneously determina¬tion of NO, NO2, O3, CO2 and H2O exchange rates. Calculations of NO, NO2 and O3 fluxes based on generally small differences (∆mi) measured between inlet and outlet of the chamber. Consequently a high accuracy and specificity of the analyzer is necessary. To achieve these requirements a highly specific NO/NO2 analyzer was used and the whole measurement system was optimized to an enduring measurement precision.rnData analysis resulted in a significant mcomp,NO2 only if statistical significance of ∆mi was detected. Consequently, significance of ∆mi was used as a data quality criterion. Photo-chemical reactions of the NO-NO2-O3 triad in the dynamic plant chamber’s volume must be considered for the determination of NO, NO2, O3 exchange rates, other¬wise deposition velocity (vdep,NO2) and mcomp,NO2 will be overestimated. No significant mcomp,NO2 for spruce could be determined under laboratory conditions, but under field conditions mcomp,NO2 could be identified between 0.17 and 0.65 ppb and vdep,NO2 between 0.07 and 0.42 mm s-1. Analyzing field data of oak, no NO2 compensation point concentration could be determined, vdep,NO2 ranged between 0.6 and 2.71 mm s-1. There is increasing indication that forests are mainly a sink for NO2 and potential NO2 emissions are low. Only when assuming high NO soil emissions, more NO2 can be formed by reaction with O3 than plants are able to take up. Under these circumstance forests can be a source for NO2.

Lightweight Internet Traffic Classification - A Subject Based Solution with Word Embeddings

Internet traffic classification is a relevant and mature research field, anyway of growing importance and with still open technical challenges, also due to the pervasive presence of Internet-connected devices into everyday life. We claim the need for innovative traffic classification solutions capable of being lightweight, of adopting a domain-based approach, of not only concentrating on application-level protocol categorization but also classifying Internet traffic by subject. To this purpose, this paper originally proposes a classification solution that leverages domain name information extracted from IPFIX summaries, DNS logs, and DHCP leases, with the possibility to be applied to any kind of traffic. Our proposed solution is based on an extension of Word2vec unsupervised learning techniques running on a specialized Apache Spark cluster. In particular, learning techniques are leveraged to generate word-embeddings from a mixed dataset composed by domain names and natural language corpuses in a lightweight way and with general applicability. The paper also reports lessons learnt from our implementation and deployment experience that demonstrates that our solution can process 5500 IPFIX summaries per second on an Apache Spark cluster with 1 slave instance in Amazon EC2 at a cost of $ 3860 year. Reported experimental results about Precision, Recall, F-Measure, Accuracy, and Cohen's Kappa show the feasibility and effectiveness of the proposal. The experiments prove that words contained in domain names do have a relation with the kind of traffic directed towards them, therefore using specifically trained word embeddings we are able to classify them in customizable categories. We also show that training word embeddings on larger natural language corpuses leads improvements in terms of precision up to 180%.

Software and Space: Investigating How a Cosmology Research Group Enacts Infrastructure by Producing Software

Thesis (Ph.D.)--University of Washington, 2016-08