Search

This paper presents a study on resonant converter topologies targeted for CMOS integration. Design methodologies to optimize efficiency for the integration of Quasi-Resonant and Quasi-Square-Wave converters are proposed. A power loss model is used to optimize the design parameters of the power stage, including the driver circuits, and also to conclude about CMOS technology limitations. Based on this discussion, and taking as reference a 0.35μm CMOS process, two converters are designed to validate the proposal: a Quasi Resonant boost converter operating at 100MHz and a Quasi-Square-Wave buck converter operating at 70MHz. Simulation results confirm the feasibility of these topologies for monolithic integration., i-ETC : ISEL Academic Journal of Electronics Telecommunications and Computers, Vol 2, No 1 (2013): The CETC2011 Issue

Submitted by Joana Lopes Dias (jdias@ubi.pt) on 2015-02-17T12:29:48Z No. of bitstreams: 1 Dissertação Vitor Costa.pdf: 756706 bytes, checksum: 4f03fe4d6cb0efde50f745f626cfd5bb (MD5) Made available in DSpace on 2015-02-17T12:29:53Z (GMT). No. of bitstreams: 1 Dissertação Vitor Costa.pdf: 756706 bytes, checksum: 4f03fe4d6cb0efde50f745f626cfd5bb (MD5) Previous issue date: 2013

LIPIcs, Volume 17, ICLP'12, Complete Volume, LIPIcs, Vol. 17, Technical Communications of the 28th International Conference on Logic Programming (ICLP'12), pages 0-0

Learning from electronic medical records (EMR) is challenging due to their relational nature and the uncertain dependence between a patient's past and future health status. Statistical relational learning is a natural fit for analyzing EMRs but is less adept at handling their inherent latent structure, such as connections between related medications or diseases. One way to capture the latent structure is via a relational clustering of objects. We propose a novel approach that, instead of pre-clustering the objects, performs a demand-driven clustering during learning. We evaluate our algorithm on three real-world tasks where the goal is to use EMRs to predict whether a patient will have an adverse reaction to a medication. We find that our approach is more accurate than performing no clustering, pre-clustering, and using expert-constructed medical heterarchies., Comment: Appears in Proceedings of the 29th International Conference on Machine Learning (ICML 2012)

O objetivo deste estudo foi determinar e comparar o limiar anaeróbio (LAn) obtido por quatro diferentes métodos com o máximo estado estável de lactato (MLSS) em corredores de endurance. Nove corredores moderadamente treinados realizaram, em diferentes dias, os seguintes testes: um teste incremental máximo para determinação do consumo máximo de oxigênio (VO2max), velocidade correspondente ao VO2max (vVO2max) e resposta do lactato sanguíneo e; dois a cinco testes de intensidade constante, com 30 min de duração, para determinação do MLSS. A partir do teste incremental, foram utilizados quatro métodos de determinação do LAn: LAn1 - velocidade correspondente a [La] fixa de 3,5 mmol.L-1; LAn2 - velocidade referente a [La] do menor equivalente [La]-velocidade somado com 1,5 mmol.L-1; LAn3 - velocidade correspondente ao Dmax; LAn4 - velocidade anterior ao segundo incremento consecutivo de [La] maior que 0,5 mmol.L-1. Não existiram diferenças significativas entre o MLSS e o LAn determinado pelos quatro métodos estudados. Entretanto, a análise de Bland-Altman expressou a extensão da discordância entre as variáveis quando os sujeitos foram analisados individualmente. Houve correlações significativas entre MLSS e LAn1 (r = 0,68; p = 0,04) e entre MLSS e o LAn2 (r = 0,79; p = 0,01). Assim, apesar de não haver diferença significativa entre os métodos de determinação do LAn com o MLSS, deve-se ter cautela para utilizá-los de forma intercambiável. The aim of this study was to compare anaerobic threshold (AT) as determined by four different methods with maximal lactate steady state (MLSS) in endurance runners. Nine moderately trained runners performed the following tests on different days: a maximal incremental exercise test to determine maximal oxygen uptake (VO2max), velocity at VO2max (vVO2max), and blood lactate response; and two to five 30-min constant load tests to determine MLSS. Based on the incremental test, four methods of AT determination were used: AT1 - velocity at 3.5 mmol.L-1 blood lactate; AT2 - velocity corresponding to the minimum lactate-velocity equivalentplus 1.5 mmol.L-1; AT3 - velocity at Dmax; and AT4 - velocity before the second consecutive blood lactate increase greater than 0.5 mmol.L-1. There were no significant differences between MLSS and AT as determined by four different methods. However, the Bland-Altman analysis showed the extent of disagreement between variables when the subjects were analyzed individually. MLSS was significantly correlated with AT1 (r=0.68; p=0.04) and AT2 (r=0.79; p=0.01). Thus, although no significant differences were found between AT methods and MLSS, one should be cautious about using these methods interchangeably.

These are the revised versions of the papers presented at CICLOPS 2011, a workshop colocated with ICLP 2011.

Statistical relational learning allows algorithms to simultaneously reason about complex structure and uncertainty with a given domain. One common challenge when analyzing these domains is the presence of latent structure within the data. We present a novel algorithm that automatically groups together different objects in a domain in order to uncover latent structure, including a hierarchy or even heterarchy. We empirically evaluate our algorithm on two large real-world tasks where the goal is to predict whether a patient will have an adverse reaction to a medication. We found that the proposed approach produced a more accurate model than the baseline approach. Furthermore, we found interesting latent structure that was deemed to be relevant and interesting by a medical collaborator. ispartof: pages:8- ispartof: Proceedings of NIPS 2011 workshop: From Statistical Genetics to Predictive Models in Personalized Medicine pages:8- ispartof: NIPS 2011 Workhop: From Statistical Genetics to Predictive Models in Personalized Medicine location:Granada, Spain date:16 Dec - 16 Dec 2011 status: published

O objetivo deste estudo foi verificar a resposta de frequência cardíaca (FC) durante as competições de "Cross-country" Olímpico (XCO). Quatorze "mountain bikers" foram separados em dois grupos: elite (n = 6; 26,5 ± 3,6 anos; 69,1 ± 2,1 kg; 174,0 ± 1,2 cm; 5,9 ± 0,9 % G; 9,0 ± 1,3 anos de treinamento) e amadores (n = 8; 25,6 ± 7,7 anos; 67,7 ± 7,0 kg; 175,5 ± 5,5 cm; 5,8 ± 2,1 % G; 8,3 ± 5,7 anos de treinamento). Os participantes foram submetidos a um teste progressivo para a identificação dos limiares metabólicos e seus respectivos valores de frequência cardíaca (FC) em cada zona de intensidade de esforço. Após intervalo mínimo de quatro dias os atletas da categoria elite foram avaliados através de monitores de FC durante a etapa brasileira da Copa do Mundo de XCO. Após 15 dias, todos os atletas foram avaliados no Campeonato Brasileiro de XCO. Os resultados indicaram que em ambas as competições, a média percentual da FC foi correspondente a 91-92 % da FCmáx. Nas competições, os atletas permaneceram durante diferentes tempos percentuais nas zonas de intensidade de esforço sendo 10,0-14,8% no domínio leve; 23,1-30,1% moderado e 55,1-66,9% intenso. Assim, este estudo apresenta que as competições de XCO são realizadas em alta intensidade, principalmente após a largada. The aim of this study was to verify and describe the intensity profile of cross-country mountain-biking races using heart rate (HR) recorded during races. Fourteen mountain bikers participated in two groups: elite (n = 6; 26.5 ± 3.6 years old; 69.1 ± 2.1 kg; 174.0 ± 1.2 cm; 5.9 ± 0.9 % BF; 9.0 ± 1.3 years of training) and amateurs (n = 8; 25.6 ± 7.7 years; 67.7 ± 7.0 kg; 175.5 ± 5.5 cm; 5.8 ± 2.1 % BF; 8.3 ± 5.7 years of training). Each cyclist was submitted to an incremental exercise test to determine the metabolic thresholds and the HR values at each threshold. After four days, only the athletes of elite category were tested during Brazilian round of XCO World Cup. Two weeks later both groups were evaluated in XCO Brazilian National Championship. The results showed that HR averaged 91-92% of HRmax. The exercise intensity zones showed different percentage of time spent during the races (10.0-14.8% easy; 23.1-30.1% moderate; 55.1-66.9% hard). This study shows that cross-country events are conducted at very high intensity, especially at the start of the race.

This research investigates how homoerotic males represent themselves in personal ads from internet. Thus, the theoretical background was Appraisal Theory, Martin & White (2005), discoursive studies from Bakhtin s Circle, Cultural Gender Studies and Social Representation Theory, Moscovici (1981). Appraisal Theory explains the evaluative aspect of language, three semantic fields (affect, appreciation and judgement). According to Bakhtin s Circle Studies, language is a social phenomena. Cultural Gender Studies there is Male Domination and Heterosexism in our social cultural context. Social representations are entities of mental, social, cultural and historical order, constructed collectively, making possible to social groups to perceive, to understand and to act on the world (MOSCOVICI, 2003). It was analyzed 308 personal ads from the following sites: Almas Gêmeas (10), Super Encontros (25), Manhunt (122), GLS Planet (81) e Viva Street (70). The analysis was based on three dimensional model to discourse from Fairclough (2001), which comprises of social practice, discursive practice and text. In text analysis, it was explored the semantic fields of Appraisal Theory (affect, judgement and appreciation), which was used to define how the advertisers represented themselves. The main categories had been: affect (romantic, promiscuous, antieffeminate, interracials, intraracials, friends, cautious and adventurers), appreciation (bears, shaved, novices, experts, olds, fits, unfits, whites and afros) and judgment (effeminates, discretes, actives, passives, versatiles and serious). Esta tese aborda como sujeitos masculinos homoeróticos se representam em anúncios pessoais eletrônicos. Para tanto, o suporte teórico utilizado foi a Teoria da Avaliatividade, de Martin & White (2006), os estudos enunciativos do Círculo de Bakhtin, os Estudos Culturais de Gênero e a Teoria das Representações Sociais, de Moscovici (1981). A Teoria da Avaliatividade, de Martin & White (2005) explica o aspecto valorativo da linguagem, partindo de três campos semânticos: afeto, apreciação e julgamento. De acordo com o Círculo de Bakhtin, a verdadeira substância da língua é o fenômeno social da interação verbal que ocorre através das enunciações. Os Estudos Culturais de Gênero afirmam que, em nosso contexto cultural, existe a dominação masculina e o heterossexismo. As representações sociais são entidades de ordem mental, social, cultural e histórica, construídas coletivamente, possibilitando aos grupos sociais perceberem, compreenderem e atuarem sobre o mundo (MOSCOVICI, 2003). Foram analisados 305 anúncios dos seguintes sites: Almas Gêmeas (10), Super Encontros (25), Manhunt (122), GLS Planet (78) e Viva Street (70). Os procedimentos de análise foram baseados no modelo tridimensional de Fairclough (2001) composto por: prática social, prática discursiva e texto. Dentro da análise da dimensão do texto, exploramos as categorias da avaliatividade (afeto, julgamento e apreciação), as quais serviram como parâmentros para identificar os modos de representações dos anunciantes. As principais categorias encontradas foram: afeto (românticos, promiscuos, antiafeminados, interraciais, intrarraciais, amigos, cautelosos e aventureiros), apreciação (ursos, lisinhos, novatos, experientes, coroas, sarados, não sarados, arianos e afros) e julgamento( afeminados, discretos, ativos, passivos, versáteis e sérios).

The free radical theory of aging posits oxidative damage to macromolecules as a primary determinant of lifespan. Recent studies challenge this theory by demonstrating that in some cases, longevity is enhanced by inactivation of oxidative stress defenses or is correlated with increased, rather than decreased reactive oxygen species and oxidative damage. Here we show that, in Saccharomyces cerevisiae, caloric restriction or inactivation of catalases extends chronological lifespan by inducing elevated levels of the reactive oxygen species hydrogen peroxide, which activate superoxide dismutases that inhibit the accumulation of superoxide anions. Increased hydrogen peroxide in catalase-deficient cells extends chronological lifespan despite parallel increases in oxidative damage. These findings establish a role for hormesis effects of hydrogen peroxide in promoting longevity that have broad implications for understanding aging and age-related diseases., This work was supported by Fundação para a Ciência e Tecnologia, Portugal Grant PTDC/AGR-ALI/71460/2006, and National Cancer Institute Cancer Center Support Grant (P30 CA016056) to Roswell Park Cancer Institute. A.M, A.S., and B.S.-M. have fellowships from Fundação para a Ciência e Tecnologia (SFRH/BD/32464/2006, SFRH/BD/33125/2007, and SFRH/BD/41674/2007, respectively).

ProbLog is a recent probabilistic extension of Prolog, where facts can be labeled with mutually independent probabilities that they belong to a randomly sampled program. The implementation of ProbLog on top of the YAP-Prolog system provides various inference algorithms that calculate the success probability of a query, i.e. the probability that the query is provable in a randomly sampled program. We discuss extensions of these algorithms with tabling that broaden the class of problems that can be handled. First, exploiting structure sharing can speed up inference in domains where different proofs of a query share many subgoals. Second, we extend exact inference to deal with negated ground subgoals in clause bodies. ispartof: International Workshop on Statistical Relational Learning location:Leuven, Belgium date:2 Jul - 4 Jul 2009 status: published

We study how probabilistic reasoning and inductive querying can be combined within ProbLog, a recent probabilistic extension of Prolog. ProbLog can be regarded as a database system that supports both probabilistic and inductive reasoning through a variety of querying mechanisms. After a short introduction to ProbLog, we provide a survey of the different types of inductive queries that ProbLog supports, and show how it can be applied to the mining of large biological networks. ispartof: CW Reports nrpages: 27 status: published

The heart rate recovery is used as indicator of aerobic condition. The objective of this study was to verify the association between markers of aerobic condition and the phases of heart rate recovery. Eleven individuals were submitted on the maximal progressive test and post-recovery was realized in passive form. The load and heart rate in deflection point of heart rate (WDMAX e HRDMAX), in maximal load (WMAX e HRMAX) and the absolute and relative values of heart rate recovery in 1, 3 and 5 minutes after exercise was computed. The significant associations were found between WMAX and relative recovery of heart rate in 3 and 5 (0.832 and 0.708; p < 0.05, respectively) and between WMAX and absolute heart rate in 3 minutes (-0.766; p < 0.05). The WDMAX was associated only with absolute values in 3 and 5 minutes (-0.809 and –0.692; p < 0.05; respectively). The best level of aerobic condition is associated with slower phase of heart rate recovery, permitting a fast decline in heart rate in this period in individuals trained. A recuperação da freqüência cardíaca (FC) é utilizada como indicador de aptidão aeróbia. O objetivo deste estudo foi associar índices de aptidão aeróbia com diversas fases da FC pós-exercício. Onze indivíduos foram submetidos a um teste progressivo máximo, com a recuperação realizada passivamente. Foram identificadas a carga e a FC do ponto de deflexão (WDMAX e FCDMAX), a FC e carga máxima (FCMAX e WDMAX) e os valores absolutos e percentuais de queda da FCREC nos minutos 1, 3 e 5 após o exercício. Foram encontradas associações entre WMAX com %FCREC 3 e 5 (0,832 e 0,708; p < 0,05; respectivamente) e com FCREC 3 (-0.766; p < 0,05). A WDMAX foi associada apenas com os valores absolutos em 3 e 5 minutos (-0,809 e –0,692; p < 0,05; respectivamente). O melhor nível de aptidão aeróbia parece estar relacionado com o comportamento da recuperação da FC na sua fase mediata, possibilitando uma queda mais rápida neste período.

The aim of this study was to compare the linear and sigmoid adjustments for the behavior of the heart rate (HR) during graded exercise test and to verify the validity of these models to identify metabolic thresholds from Boltzman's non-linear adjustment during incremental exercise. Ten high level mountain bikers (27.1 ± 7.4 years; 67.0 ± 6.1 kg; 175.1 ± 4.1 cm; 6.1 ± 1.9% estimated body fat) were submitted to graded exercise test in cycle-simulator, with initial load of 100 W and 30 W increments every 3 min until exhaustion. The HR behavior was fitted by Boltzman's sigmoid equation, and HR inflection point (HRIP = fixed value of V50 from equation) and HR deflection point (HRDP = Dmax method) were identified. Blood lactate samples were withdrawn at the end of each stage. The first lactate threshold (LT1) was determined by the calculation of minimum value equivalent of lactate/work. To identify the second lactate threshold (LT2), 1.5 mmol.l-1 was added to the LT1 value. In all athletes the sigmoid model was more appropriate than the linear one (p < 0.0001). The HRIP (161 ± 10 bpm) and HRDP (179 ± 5 bpm) were identified in 8 and 7 subjects, respectively. From the similarity of the results about the intensity in power output and HR, it can be suppose that the HRIP is in the same physiological domain of the LT2, above to LT1 and lower to HRDP. However, were found poor association between lactate thresholds, HRIP and HRDP (-0.08 to 0.37). O objetivo deste estudo foi comparar os ajustes linear e sigmóide da freqüência cardíaca (FC) durante o exercício incremental e analisar a validade destes modelos para identificar os limiares de transição metabólica em exercício progressivo, a partir de pontos de mudança no ajuste não-linear de Boltzmann. Dez "mountain bikers" de alto rendimento (27,1 ± 7,4 anos; 67,0 ± 6,1 kg; 175,1 ± 4,1 cm; 6,1 ± 1,9 % de gordura) foram submetidos a teste progressivo no ciclo-simulador, com carga inicial de 100 W e incremento de 30 W a cada 3 min até a exaustão. A FC foi registrada batimento-a-batimento durante todo o teste. A curva da FC foi ajustada pela equação sigmóide de Boltzmann, com cálculo do Ponto de Inflexão da FC (PIFC = valor fixo V50 da equação) e Ponto de Deflexão da FC (PDFC = método Dmáx). No final de cada estágio, foi determinada a concentração sanguínea de lactato ([La]). O primeiro limiar de lactato (LL1) foi determinado através do cálculo do menor valor equivalente de lactato/carga. Para o segundo limiar (LL2), foi acrescido 1,5 mmol.l-1 ao menor valor deste equivalente. Em todos os atletas, o ajuste sigmóide da FC se apresentou mais adequado que o linear (p < 0,0001). O PIFC (161 ± 10 bpm) e PDFC (179 ± 5 bpm) foram identificados em oito e sete sujeitos, respectivamente. A FC e a carga do PIFC foram semelhantes às obtidas no LL2, enquanto que esses parâmetros foram menores no LL1 e maiores no PDFC. Assim, é possível especular que o PIFC está no mesmo domínio fisiológico do LL2. Entretanto, foram encontradas fracas associações entre os limiares de lactato com o PIFC e PDFC (-0,08 a 0,37).

ispartof: pages:117-131 ispartof: Proceedings of CICLOPS 2007, 7th International Colloquium on Implementation of Constraint and LOgic Programming Systems pages:117-131 ispartof: 7th International Colloquium on Implementation of Constraint and Logic Programming Systems location:Porto, Portugal date:8 Sep - 8 Sep 2007 status: published

Motivation: Profile hidden Markov Models (pHMMs) are a popular and very useful tool in the detection of the remote homologue protein families. Unfortunately, their performance is not always satisfactory when proteins are in the 'twilight zone'. We present HMMER-STRUCT, a model construction algorithm and tool that tries to improve pHMM performance by using structural information while training pHMMs. As a first step, HMMER-STRUCT constructs a set of pHMMs. Each pHMM is constructed by weighting each residue in an aligned protein according to a specific structural property of the residue. Properties used were primary, secondary and tertiary structures, accessibility and packing. HMMER-STRUCT then prioritizes the results by voting. Results: We used the SCOP database to perform our experiments. Throughout, we apply leave-one-family-out cross-validation over protein superfamilies. First, we used the MAMMOTH-mult structural aligner to align the training set proteins. Then, we performed two sets of experiments. In a first experiment, we compared structure weighted models against standard pHMMs and against each other. In a second experiment, we compared the voting model against individual pHMMs. We compare method performance through ROC curves and through Precision/Recall curves, and assess significance through the paired two tailed t-test. Our results show significant performance improvements of all structurally weighted models over default HMMER, and a significant improvement in sensitivity of the combined models over both the original model and the structurally weighted models., Comment: 6 pages, 7 figures

OBJETIVO: Investigar em gestantes possível correlação da infecção pelo Helicobacter pylori com sintomas dispépticos e características maternas da gravidez. MÉTODOS: A soropositividade ao H. pylori foi obtida em 146 mulheres com 10 a 23 semanas de gestação. A média de idade foi de 29,6 anos, variando de 17 a 45 anos. A determinação semiquantitativa de anticorpos da classe IgG contra o H. pylori foi realizada com "kit" Elisa de procedência comercial (Varelisa H.pylori IgG Antibodies, Pharmacia & Upjohn Diagnostics GmbH & Co., Germany). Grávidas com sintomas dispépticos que necessitaram de tratamento constituíram o grupo investigado. Outras características maternas também foram correlacionadas à infecção pelo H. pylori: obesidade - IMC >30; paridade - primípara e multípara; evolução da gestação - a termo e aborto ou prematuridade. Os valores obtidos foram submetidos a análise estatística - t de Student e Qui-quadrado. RESULTADOS: A prevalência da infecção pelo H. pylori na população estudada foi de 65,7%. Constatou-se ausência de correlação entre a infecção bacteriana e os fenômenos dispépticos da gravidez. Correlação negativa também foi constatada para obesidade, paridade e o evoluir da gestação. Gestantes com sorologia positiva apresentaram media de idade significantemente maior que as de sorologia negativa: 30,45 ± 6,87 versus 27,96 ± 7,44, p= 0,045565. CONCLUSÃO: O presente estudo não associou a infecção pelo H. pylori com os sintomas dispépticos da gravidez, nem com as características maternas estudadas: obesidade, paridade e evolução da gravidez. OBJECTIVE: The present study intended to investigate whether there was a relationship between pregnant women with dyspeptic complaints and with and without Helycobacter pylori infection in the first and second trimester of pregnancy. METHODS: H.pylori seropositivity was determined in 146 (aged 20 to 40 years) antenatal patients at 10 to 23 weeks gestation. H.Pylori ser um Immunoglobulin IgG antibody concentrations were determined by enzyme-linked immunoadsorbent assay (ELISA). Age, obesity, parity, outcome of pregnancy and dyspeptic symptoms were registered in the patients' medical records. Chi-square and Student's t-test were used for statistical analysis of the data. RESULTS: Prevalence of H.pylori infection in the population studied was 65.7% (96 of 146). H pylori infection was not associated with parity, preterm delivery, obesity or with dyspeptic symptoms. Women positive for H.pylori IgG were older (30, 45 ± 6.87 vs 27, 96 ± 7.44, p= 0,045565*) than H. pylori negative patients. CONCLUSION: In this study there was no association of H pylori infection with dyspeptic symptoms in pregnancy .nor was its presence associated with obesity, parity and outcome of pregnancy.

Several techniques for implementing Prolog in a efficient manner have been devised since the original interpreter, many of them aimed at achieving more speed. There are two main approaches to efficient Prolog implementation: (1) compilers to bytecode and then interpreting it (emulators) or (2) compilers to native code. Emulators have smaller load/compilation time and are a good solution for their simplicity when speed is not a priority. Compilers are more complex than emulators, and the difference is much more acute if some form of code analysis is performed as part of the compilation, which impacts development time. Generation of low level code promises faster programs at the expense of using more resources during the compilation phase. In our work besides using an mixed execution mode, we design an optimizing compiler that using type feedback profiling, dynamic compilation and dynamic deoptimization for improving the performance of logic programming languages.

The present study intended to investigate whether there was a relationship between pregnant women with dyspeptic complaints and with and without Helicobacter pylori infection in the first and second trimester of pregnancy.H.pylori seropositivity was determined in 146 (aged 20 to 40 years) antenatal patients at 10 to 23 weeks gestation. H.Pylori ser um Immunoglobulin IgG antibody concentrations were determined by enzyme-linked immunoadsorbent assay (ELISA). Age, obesity, parity, outcome of pregnancy and dyspeptic symptoms were registered in the patients' medical records. Chi-square and Student's t-test were used for statistical analysis of the data.Prevalence of H.pylori infection in the population studied was 65.7% (96 of 146). H pylori infection was not associated with parity, preterm delivery, obesity or with dyspeptic symptoms. Women positive for H.pylori IgG were older (30, 45 +/- 6.87 vs 27, 96 +/- 7.44, p= 0.045565*) than H. pylori negative patients.In this study there was no association of H pylori infection with dyspeptic symptoms in pregnancy, nor was its presence associated with obesity, parity and outcome of pregnancy.

Identity Equivalence or Alias Detection is an important topic in Intelligence Analysis. Often, terrorists will use multiple different identities to avoid detection. We apply machine learning to the task of determining Identity Equivalence. Two challenges exist in this domain. First, data can be spread across multiple tables. Second, we need to limit the number of false positives. We present a two step approach to combat these issues. First, we use Inductive Logic Programming to find a set of rules that are predictive of aliases. In the second step, we treat each learned rule as a random variable in a Bayesian Network. We use the Bayesian Network to assign a probability that two identities are aliases. We evaluate our technique on several data sets and find that layering Bayesian Network over the rules significantly increases the precision of our system. ispartof: Proceedings of the International Conference on Intelligence Analysis ispartof: Proceedings of the International Conference on Intelligence Analysis location:McLean, VA, USA date:2 May - 6 May 2005 status: published

Interpreted languages are widely used due to ease to use, portability, and safety. On the other hand, interpretation imposes a significance overhead. Just­in­ Time (JIT) compilation is a popular approach to improving the runtime performance of languages such as Java. We compare the performance of a JIT compiler with a traditional compiler and with an emulator. We show that the compilation overhead from using JIT is negligible, and that the JIT compiler achieves better overall performance, suggesting the case for aggresive compilation in JIT compilers.