Taiane Schaedler Prass

In this note we develop a new technique for parameter estimation of univariate time series by means of a parametric copula approach. The proposed methodology is based on a relationship between a process' covariance decay and parametric bivariate copulas associated to lagged variables. This relationship provides a way for estimating parameters that are identifiable through the process' covariance decay, such as in long range dependent processes. We provide a rigorous asymptotic theory for the proposed estimator. We also present a Monte Carlo simulation study to asses the finite sample performance of the proposed estimator.

Abstract In this work, we introduce a class of dynamic models for time series taking values on the unit interval. The proposed model follows a generalized linear model approach where the random component, conditioned on the past information, follows a beta distribution, while the conditional mean specification may include covariates and also an extra additive term given by the iteration of a map that can present chaotic behavior. The resulting model is very flexible and its systematic component can accommodate short- and long-range dependence, periodic behavior, laminar phases, etc. We derive easily verifiable conditions for the stationarity of the proposed model, as well as conditions for the law of large numbers and a Birkhoff-type theorem to hold. A Monte Carlo simulation study is performed to assess the finite sample behavior of the partial maximum likelihood approach for parameter estimation in the proposed model. Finally, an application to the proportion of stored hydroelectrical energy in Southern Brazil is presented.

In this work, we develop the asymptotic theory of the Detrended Fluctuation Analysis (DFA) and Detrended Cross-Correlation Analysis (DCCA) for trend-stationary stochastic processes without any assumption on the specific form of the underlying distribution. All results are presented and derived under the general framework of potentially overlapping boxes for the polynomial fit. We prove the stationarity of the DFA and DCCA, viewed as stochastic processes, obtain closed forms for moments up to second order, including the covariance structure for DFA and DCCA and a miscellany of law of large number related results. Our results generalize and improve several results presented in the literature. To verify the behavior of our theoretical results in small samples, we present a Monte Carlo simulation study and an empirical application to econometric time series.

Power plant performance can decrease along with its life span, and move away from the design and commissioning targets. Maintenance issues, operational practices, market restrictions, and financial objectives may lead to that behavior, and the knowledge of appropriate actions could support the system to retake its original operational performance. This paper applies unsupervised machine learning techniques to identify operating patterns based on the power plant’s historical data which leads to the identification of appropriate steam generator efficiency conditions. The selected operational variables are evaluated in respect to their impact on the system performance, quantified by the Variable Importance Index. That metric is proposed to identify the variables among a much wide set of monitored data whose variation impacts the overall power plant operation, and should be controlled with more attention. Principal Component Analysis (PCA) and k-means++ clustering techniques are used to identify suitable operational conditions from a one-year-long data set with 27 recorded variables from a steam generator of a 360MW thermal power plant. The adequate number of clusters is identified by the average Silhouette coefficient and the Variable Importance Index sorts nine variables as the most relevant ones, to finally group recommended settings to achieve the target conditions. Results show performance gains in respect to the average historical values of 73.5% and the lowest efficiency condition records of 68%, to the target steam generator efficiency of 76%.

Oral and oropharyngeal cancer is considered a public health problem in several countries due to its high incidence and mortality rate.

Objective:

This study aimed to analyze oral and oropharyngeal cancer mortality in Uruguay from 1997 to 2014 by age, sex and country region.

Methodology:

A time series ecological study using secondary data was performed. Data on mortality due to oral and oropharyngeal cancers were obtained from the Vital Statistics Department of Uruguay's Ministry of Public Health.

Results:

The cumulative mortality rate due to oral and oropharyngeal cancer over the study period was of 19.26/100,000 persons in women and 83.61/100.000 in men, with a mean annual rate of 1.75/100,000 in women and 7.60/100,000 in men. Mortality rate from both sites during the study period was 4.34 times higher in men than in women. Malignant neoplasms of other parts of the tongue and base of tongue showed the highest mortality rate. The means of the annual coefficients of deaths were higher for the age groups between 50 and 69 years. Higher mortality rates of oral and oropharyngeal cancer were observed in Artigas (4.63) and Cerro Largo (3.75).

Conclusions:

Our study described a high mortality rate for oral and oropharyngeal cancer in Uruguay from 1997 to 2014. According to the country's health department, men, tongue cancer, and oral cavity had higher mortality rates, with some variation. Prevention strategies with control of risk factors and early diagnosis are necessary to improve survival in the Uruguayan population.

Generalized observations of temporal trends in mortality could mask relevant specific patterns. The objective of this study is to analyze the trend of oral and oropharyngeal cancer mortality rates in Brazil,2000 to 2013, considering the differences by sex, anatomical site, age group and race. Data on oral and oropharyngeal cancer mortality were obtainedthe Mortality Information System. The trend of historical series mortality rates, by stratum, was estimated by generalized linear regression using the Prais-Winsten method. From 2000 to 2013, there were 61,190 deathsoral cancer and oropharynx (average of annual coefficients: 3.50 deaths/100 thousand inhabitants). The trend of mortality rates was stationary for males and increasing for females (1.31%/year). Growth pattern for men aged 20-29 years (2.92%/year) and for men of race/color brown (20.36%/year). Growth pattern was also identified for white women (2.70%/year) and brown women (8.24%/year). The surveillance of this condition should consider the sociodemographic differences of the population for an equitable planning of the strategies of care, because they reflected in different trends of mortality rates by oral and oropharynx cancer in Brazil.

Abstract
BACKGROUND:
To analyze the trends of oral and oropharyngeal cancer mortality in Uruguay between 1997 and 2014 according to sex and age groups and its possible association with sociodemographic factors.

MATERIAL AND METHODS:
A time-series ecological study using secondary data was performed. The data about mortality due to oral and oropharyngeal cancers were obtained from the Statistics Vitals Department of the Public Health Ministry of Uruguay. To estimate the mortality trends of the historical series, by sex, anatomical site and age groups, linear regressions generated by the Prais-Winsten procedure were used.

RESULTS:
The analysis of mortality trends for oral cavity and oropharyngeal cancers in Uruguay indicated that the global mortality rate was stable over the studied period. The women's mortality rate increased from 0.51 per 100,000 in 1997 to 0.65 per 100,000 in 2014 while for men, rates per 100,000 went from 3.22 in 1997 to 2.20 per 100,000 in 2014. Mortality from oral cancer in men decreased between 1997 and 2014. Mortality by oropharyngeal cancer, irrespective of sex, remained stable. Analysis by cancer site revealed decreasing trends tumors situated in the base of the tongue and gum. Years of education, unemployment, smoking and Gini index were not associated with mortality trends.

CONCLUSIONS:
The overall mortality from oral and oropharyngeal cancer in Uruguay has remained constant in the period between 1997 and 2014. Oral cancer mortality decreased in men and increased in women and decreased at the base of the tongue. It's necessary to continue monitoring the behavior of these diseases.

The objective was to investigate if there is an association between the mortality rates due to oral and oropharyngeal cancer in Brazil and the expansion of access to public primary and specialized dental care services that resulted from the implementation of the National Oral Health Policy, between 2000 and 2013. The mortality data were obtained from the records of the Mortality Information System and the exposure variables were obtained from databases of the Brazilian Ministry of Health and the Brazilian Institute of Geography and Statistics. The main exposures investigated were “coverage of primary dental care” and “number of specialized dental care centers”. Additional covariates included “Gini index of household income”, “average number of years of study”, “proportion of unemployed people” and “proportion of smokers”. For the statistical analysis, a random coefficient model was used. There was a statistically significant association between the mortality rates by oral and oropharyngeal cancer with coverage by primary dental care and the number of specialized dental care centers with males. This study found that the expansion of the coverage of primary dental care and the number of specialized dental care centers are associated with the reduction of mortality rates due to oral and oropharyngeal cancer in Brazil. There is plausibility for the association found, which needs to be confirmed by implementation studies.

In this work we introduce the class of beta autoregressive fractionally integrated moving average models for continuous random variables taking values in the continuous unit interval (0,1). The proposed model accommodates a set of regressors and a long-range dependent time series structure. We derive the partial likelihood estimator for the parameters of the proposed model, obtain the associated score vector and Fisher information matrix. We also prove the consistency and asymptotic normality of the estimator under mild conditions. Hypotheses testing, diagnostic tools and forecasting are also proposed. A Monte Carlo simulation is considered to evaluate the finite sample performance of the partial likelihood estimators and to study some of the proposed tests. An empirical application is also presented and discussed.

Objective: Vaccines are effective in controlling and eradicating infectious diseases. However, adverse events following immunization (AEFI) can occur in susceptible individuals. The objective of this study was to analyze the Brazilian AEFI database and compare eight vaccines in order to profile risks of AEFIs related to the mandated pediatric schedule of immunization, considering the age and sex of the child, type of vaccine, and reported adverse events. Methods: We analyzed the Brazilian AEFI database integrating reports between 2005 and 2010 for children less than 10-years old immunized with eight mandated vaccines: diphtheria, pertussis, tetanus, Haemophilus influenzae type b (TETRA); diphtheria, tetanus, and pertussis (DTP); Bacillus Calmette–Guerin (BCG); oral poliovirus vaccine (OPV); measles, mumps, and rubella (MMR); oral rotavirus vaccine (ORV); hepatitis B (HB); and yellow fever (YF). We compared the children’s age regarding types of AEFI, evaluated AEFI factors associated with the chance of hospitalization of the child, and estimated the chance of notification of an AEFI as a function of the type of vaccine. In total, 47,105 AEFIs were observed for the mandated vaccines. Results: The highest AEFI rate was for the TETRA vaccine and the lowest was for the OPV vaccine, with 60.1 and 2.3 events per 100,000 inoculations, respectively. The TETRA vaccine showed the highest rate of hypotonic hyporesponsive episode, followed by convulsion and fever. The MMR and YF vaccines were associated with generalized rash. BCG was associated with enlarged lymph glands but showed the largest negative (protective) association with hyporesponsive events and seizures. Compared with children aged 5–9-years old, young children (<1 year) showed significantly higher odds of hospitalization. Conclusions: The Brazilian AEFI registry is useful to compare the magnitude and certain characteristics of adverse events associated with mandated pediatric vaccines.

Bayesian inference for fractionally integrated exponential generalized autoregressive conditional heteroscedastic (FIEGARCH) models using Markov chain Monte Carlo (MCMC) methods is described. A simulation study is presented to assess the performance of the procedure, under the presence of long-memory in the volatility. Samples from FIEGARCH processes are obtained upon considering the generalized error distribution (GED) for the innovation process. Different values for the tail-thickness parameter ν are considered covering both scenarios, innovation processes with lighter (ν > 2) and heavier (ν < 2) tails than the Gaussian distribution (ν = 2). A comparison between the performance of quasi-maximum likelihood (QML) and MCMC procedures is also discussed. An application of the MCMC procedure to estimate the parameters of a FIEGARCH model for the daily log-returns of the S&P500 U.S. stock market index is provided.

Here we present a theoretical study on the main properties of Fractionally Integrated Exponential Generalized Autoregressive Conditional Heteroskedastic (FIEGARCH) processes. We analyze the conditions for the existence, the invertibility, the stationarity and the ergodicity of these processes. We prove that, if {Xt}t∈Z is a FIEGARCH(p,d,q) process then, under mild conditions, {ln(Xt2)}t∈Z is an ARFIMA(q,d,0) with correlated innovations, that is, an autoregressive fractionally integrated moving average process. The convergence order for the polynomial coefficients that describes the volatility is presented and results related to the spectral representation and to the covariance structure of both processes {ln(Xt2)}t∈Z and {ln(σt2)}t∈Z are discussed. Expressions for the kurtosis and the asymmetry measures for any stationary FIEGARCH(p,d,q) process are also derived. The h-step ahead forecast for the processes {Xt}t∈Z, {ln(σt2)}t∈Z and {ln(Xt2)}t∈Z are given with their respective mean square error of forecast. The work also presents a Monte Carlo simulation study showing how to generate, estimate and forecast based on six different FIEGARCH models. The forecasting performance of six models belonging to the class of autoregressive conditional heteroskedastic models (namely, ARCH-type models) and radial basis models is compared through an empirical application to Brazilian stock market exchange index.

We consider the Fractionally Integrated Exponential Generalized Autoregressive Conditional Heteroskedasticity process, denoted by FIEGARCH(p,d,q), introduced by Bollerslev and Mikkelsen (1996). We present a simulated study regarding the estimation of the risk measure VaRp on FIEGARCH processes. We consider the distribution function of the portfolio log-returns (univariate case) and the multivariate distribution function of the risk-factor changes (multivariate case). We also compare the performance of the risk measures VaRp, ESp and MaxLoss for a portfolio composed by stocks of four Brazilian companies.

Here we develop the theory of seasonal FIEGARCH processes, denoted by SFIEGARCH, establishing conditions for the existence, the invertibility, the stationarity and the ergodicity of these processes. We analyze their asymptotic dependence structure by means of the autocovariance and autocorrelation functions. We also present some properties regarding their spectral representation. All properties are illustrated through graphical examples and an application of SFIEGARCH models to describe the volatility of the S&P500 US stock index log-return time series in the period from December 13, 2004 to October 10, 2009 is provided.

Birth weight data (22,012 live-births) from a public hospital in Porto Velho (Amazon) was used in multiple statistical models to assess the effects of forest-fire smoke on human reproductive outcome. Mean birth weights for girls (3,139 g) and boys (3,393 g) were considered statistically different (p-value < 2.2e-16). Among all models analyzed, the means were considered statistically different only when treated as a function of month and year (p-value = 0.0989, girls and 0.0079, boys) . The R2 statistics indicate that the regression models considered are able to explain 65 {%} (girls) and 54 {%} (boys) of the variation of the mean birth weight.

The paper compares forecasts of mean monthly water levels up to six months ahead at Ladário, on the Upper Paraguay River, Brazil, estimated from two long-range dependence models. In one of them, the marked seasonal cycle was removed and a fractionally differenced model was fitted to the transformed series. In the other, a seasonal fractionally differenced model was fitted to water levels without transformation. Forecasts from both models for periods up to six months ahead were compared with forecasts given by simpler “short-range dependence” Box-Jenkins models, one fitted to the transformed series, the other a seasonal autoregressive moving average (ARMA) model. Estimates of parameters in the four models (two “long-range dependence”, two “short-range dependence”) were updated at six-monthly intervals over a 20 year period, and forecasts were compared using root mean square errors (rmse) between water-level forecasts and observed levels. As judged by rmse, performances of the two long-range dependence models, and of the ARMA (1,1) short-range dependence model, were very similar; all three out-performed the seasonal short-range dependence ARMA model. There was evidence that all models performed better during recession periods, than on the hydrograph rising limb.