Abstract: The forensic use of handwriting depends on the analysis, comparison, and evaluation decisions made by forensic document examiners. When using biometric technology in forensic applications, it is necessary to compute Likelihood Ratio (LR) for quantifying strength of evidence under two competing hypotheses, namely the prosecution and the defense hypotheses wherein a set of assumptions and methods for a given data set will be made. It is therefore important to know how repeatable and reproducible our estimated LR is. This paper evaluated the accuracy and reproducibility of examiners' decisions. Confidence interval for the estimated LR were presented so as not get an incorrect estimate that will be used to deliver wrong judgment in the court of Law. The estimate of LR is fundamentally a Bayesian concept and we used two LR estimators, namely Logistic Regression (LoR) and Kernel Density Estimator (KDE) for this paper. The repeatability evaluation was carried out by retesting the initial experiment after an interval of six months to observe whether examiners would repeat their decisions for the estimated LR. The experimental results, which are based on handwriting dataset, show that LR has different confidence intervals which therefore implies that LR cannot be estimated with the same certainty everywhere. Though the LoR performed better than the KDE when tested using the same dataset, the two LR estimators investigated showed a consistent region in which LR value can be estimated confidently. These two findings advance our understanding of LR when used in computing the strength of evidence in handwriting using forensics.
Abstract: Developing countries are nowadays confronted with great challenges related to domestic sanitation services in view of the imminent water scarcity. Contemporary sanitation technologies established in these countries are likely to pose health risks unless waste management standards are followed properly. This paper provides a solution to sustainable sanitation with the development of an innovative toilet system, called Nano Membrane Toilet (NMT), which has been developed by Cranfield University and sponsored by the Bill & Melinda Gates Foundation. The particular technology converts human faeces into energy through gasification and provides treated wastewater from urine through membrane filtration. In order to evaluate the environmental profile of the NMT system, a deterministic life cycle assessment (LCA) has been conducted in SimaPro software employing the Ecoinvent v3.3 database. The particular study has determined the most contributory factors to the environmental footprint of the NMT system. However, as sensitivity analysis has identified certain critical operating parameters for the robustness of the LCA results, adopting a stochastic approach to the Life Cycle Inventory (LCI) will comprehensively capture the input data uncertainty and enhance the credibility of the LCA outcome. For that purpose, Monte Carlo simulations, in combination with an artificial neural network (ANN) model, have been conducted for the input parameters of raw material, produced electricity, NOX emissions, amount of ash and transportation of fertilizer. The given analysis has provided the distribution and the confidence intervals of the selected impact categories and, in turn, more credible conclusions are drawn on the respective LCIA (Life Cycle Impact Assessment) profile of NMT system. Last but not least, the specific study will also yield essential insights into the methodological framework that can be adopted in the environmental impact assessment of other complex engineering systems subject to a high level of input data uncertainty.
Abstract: Background: Traditional chronic disease management did not pay attention to effects of geographic factors on the compliance of treatment regime, which resulted in geographic inequality in outcomes of chronic disease management. This study aims to examine the geographic distribution and clustering of quality indicators of diabetes care. Method: We first extracted address, demographic information and quality of care indicators (number of visits, complications, prescription and laboratory records) of patients with diabetes for 2014 from medical information system in a medical center in Tainan City, Taiwan, and the patients’ addresses were transformed into district- and village-level data. We then compared the differences of geographic distribution and clustering of quality of care indicators between districts and villages. Despite the descriptive results, rate ratios and 95% confidence intervals (CI) were estimated for indices of care in order to compare the quality of diabetes care among different areas. Results: A total of 23,588 patients with diabetes were extracted from the hospital data system; whereas 12,716 patients’ information and medical records were included to the following analysis. More than half of the subjects in this study were male and between 60-79 years old. Furthermore, the quality of diabetes care did indeed vary by geographical levels. Thru the smaller level, we could point out clustered areas more specifically. Fuguo Village (of Yongkang District) and Zhiyi Village (of Sinhua District) were found to be “hotspots” for nephropathy and cerebrovascular disease; while Wangliau Village and Erwang Village (of Yongkang District) would be “coldspots” for lowest proportion of ≥80% compliance to blood lipids examination. On the other hand, Yuping Village (in Anping District) was the area with the lowest proportion of ≥80% compliance to all laboratory examination. Conclusion: In spite of examining the geographic distribution, calculating rate ratios and their 95% CI could also be a useful and consistent method to test the association. This information is useful for health planners, diabetes case managers and other affiliate practitioners to organize care resources to the areas most needed.
Abstract: This paper presents the confidence intervals for the
effect size base on bootstrap resampling method. The meta-analytic
confidence interval for effect size is proposed that are easy to
compute. A Monte Carlo simulation study was conducted to compare
the performance of the proposed confidence intervals with the
existing confidence intervals. The best confidence interval method
will have a coverage probability close to 0.95. Simulation results
have shown that our proposed confidence intervals perform well in
terms of coverage probability and expected length.
Abstract: The paper deals with the maximum likelihood estimation of the parameters of the Burr type V distribution based on left censored samples. The maximum likelihood estimators (MLE) of the parameters have been derived and the Fisher information matrix for the parameters of the said distribution has been obtained explicitly. The confidence intervals for the parameters have also been discussed. A simulation study has been conducted to investigate the performance of the point and interval estimates.
Abstract: In this paper we proposed the new confidence interval for the normal population mean with known coefficient of variation. In practice, this situation occurs normally in environment and agriculture sciences where we know the standard deviation is proportional to the mean. As a result, the coefficient of variation of is known. We propose the new confidence interval based on the recent work of Khan [3] and this new confidence interval will compare with our previous work, see, e.g. Niwitpong [5]. We derive analytic expressions for the coverage probability and the expected length of each confidence interval. A numerical method will be used to assess the performance of these intervals based on their expected lengths.
Abstract: In many practical applications in various areas, such as engineering, science and social science, it is known that there exist bounds on the values of unknown parameters. For example, values of some measurements for controlling machines in an industrial process, weight or height of subjects, blood pressures of patients and retirement ages of public servants. When interval estimation is considered in a situation where the parameter to be estimated is bounded, it has been argued that the classical Neyman procedure for setting confidence intervals is unsatisfactory. This is due to the fact that the information regarding the restriction is simply ignored. It is, therefore, of significant interest to construct confidence intervals for the parameters that include the additional information on parameter values being bounded to enhance the accuracy of the interval estimation. Therefore in this paper, we propose a new confidence interval for the coefficient of variance where the population mean and standard deviation are bounded. The proposed interval is evaluated in terms of coverage probability and expected length via Monte Carlo simulation.
Abstract: In this paper, we propose two new confidence intervals for the inverse of a normal mean with a known coefficient of variation. One of new confidence intervals for the inverse of a normal mean with a known coefficient of variation is constructed based on the pivotal statistic Z where Z is a standard normal distribution and another confidence interval is constructed based on the generalized confidence interval, presented by Weerahandi. We examine the performance of these confidence intervals in terms of coverage probabilities and average lengths via Monte Carlo simulation.
Abstract: Motivated by the recent work of Herbert, Hayen, Macaskill and Walter [Interval estimation for the difference of two independent variances. Communications in Statistics, Simulation and Computation, 40: 744-758, 2011.], we investigate, in this paper, new confidence intervals for the difference between two normal population variances based on the generalized confidence interval of Weerahandi [Generalized Confidence Intervals. Journal of the American Statistical Association, 88(423): 899-905, 1993.] and the closed form method of variance estimation of Zou, Huo and Taleban [Simple confidence intervals for lognormal means and their differences with environmental applications. Environmetrics 20: 172-180, 2009]. Monte Carlo simulation results indicate that our proposed confidence intervals give a better coverage probability than that of the existing confidence interval. Also two new confidence intervals perform similarly based on their coverage probabilities and their average length widths.
Abstract: The log periodogram regression is widely used in empirical
applications because of its simplicity, since only a least squares
regression is required to estimate the memory parameter, d, its good
asymptotic properties and its robustness to misspecification of the
short term behavior of the series. However, the asymptotic distribution
is a poor approximation of the (unknown) finite sample distribution
if the sample size is small. Here the finite sample performance of different
nonparametric residual bootstrap procedures is analyzed when
applied to construct confidence intervals. In particular, in addition to
the basic residual bootstrap, the local and block bootstrap that might
adequately replicate the structure that may arise in the errors of the
regression are considered when the series shows weak dependence in
addition to the long memory component. Bias correcting bootstrap
to adjust the bias caused by that structure is also considered. Finally,
the performance of the bootstrap in log periodogram regression based
confidence intervals is assessed in different type of models and how
its performance changes as sample size increases.
Abstract: Using a set of confidence intervals, we develop a
common approach, to construct a fuzzy set as an estimator for
unknown parameters in statistical models. We investigate a method
to derive the explicit and unique membership function of such fuzzy
estimators. The proposed method has been used to derive the fuzzy
estimators of the parameters of a Normal distribution and some
functions of parameters of two Normal distributions, as well as the
parameters of the Exponential and Poisson distributions.
Abstract: Given a bivariate normal sample of correlated variables,
(Xi, Yi), i = 1, . . . , n, an alternative estimator of Pearson’s correlation
coefficient is obtained in terms of the ranges, |Xi − Yi|.
An approximate confidence interval for ρX,Y is then derived, and
a simulation study reveals that the resulting coverage probabilities
are in close agreement with the set confidence levels. As well, a
new approximant is provided for the density function of R, the
sample correlation coefficient. A mixture involving the proposed
approximate density of R, denoted by hR(r), and a density function
determined from a known approximation due to R. A. Fisher is shown
to accurately approximate the distribution of R. Finally, nearly exact
density approximants are obtained on adjusting hR(r) by a 7th degree
polynomial.
Abstract: In this paper we proposed two new confidence intervals for the normal population mean with known coefficient of variation. This situation occurs normally in environment and agriculture experiments where the scientist knows the coefficient of variation of their experiments. We propose two new confidence intervals for this problem based on the recent work of Searls [5] and the new method proposed in this paper for the first time. We derive analytic expressions for the coverage probability and the expected length of each confidence interval. Monte Carlo simulation will be used to assess the performance of these intervals based on their expected lengths.
Abstract: Zero inflated Strict Arcsine model is a newly developed model which is found to be appropriate in modeling overdispersed count data. In this study, maximum likelihood estimation method is used in estimating the parameters for zero inflated strict arcsine model. Bootstrapping is then employed to compute the confidence intervals for the estimated parameters.
Abstract: Oxidative stress is considered to be the cause for onset
and the progression of type 2 diabetes mellitus (T2DM) and
complications including neuropathy. It is a deleterious process that
can be an important mediator of damage to cell structures: protein,
lipids and DNA. Data suggest that in patients with diabetes and
diabetic neuropathy DNA repair is impaired, which prevents effective
removal of lesions. Objective: The aim of our study was to evaluate
the association of the hOGG1 (326 Ser/Cys) and XRCC1 (194
Arg/Trp, 399 Arg/Gln) gene polymorphisms whose protein is
involved in the BER pathway with DNA repair efficiency in patients
with diabetes type 2 and diabetic neuropathy compared to the healthy
subjects. Genotypes were determined by PCR-RFLP analysis in 385
subjects, including 117 with type 2 diabetes, 56 with diabetic
neuropathy and 212 with normal glucose metabolism. The
polymorphisms studied include codon 326 of hOGG1 and 194, 399
of XRCC1 in the base excision repair (BER) genes. Comet assay was
carried out using peripheral blood lymphocytes from the patients and
controls. This test enabled the evaluation of DNA damage in cells
exposed to hydrogen peroxide alone and in the combination with the
endonuclease III (Nth). The results of the analysis of polymorphism
were statistically examination by calculating the odds ratio (OR) and
their 95% confidence intervals (95% CI) using the ¤ç2-tests. Our data
indicate that patients with diabetes mellitus type 2 (including those
with neuropathy) had higher frequencies of the XRCC1 399Arg/Gln
polymorphism in homozygote (GG) (OR: 1.85 [95% CI: 1.07-3.22],
P=0.3) and also increased frequency of 399Gln (G) allele (OR: 1.38
[95% CI: 1.03-1.83], P=0.3). No relation to other polymorphisms
with increased risk of diabetes or diabetic neuropathy. In T2DM
patients complicated by neuropathy, there was less efficient repair of
oxidative DNA damage induced by hydrogen peroxide in both the
presence and absence of the Nth enzyme. The results of our study
suggest that the XRCC1 399 Arg/Gln polymorphism is a significant
risk factor of T2DM in Polish population. Obtained data suggest a
decreased efficiency of DNA repair in cells from patients with
diabetes and neuropathy may be associated with oxidative stress.
Additionally, patients with neuropathy are characterized by even
greater sensitivity to oxidative damage than patients with diabetes,
which suggests participation of free radicals in the pathogenesis of
neuropathy.
Abstract: In this paper some procedures for building confidence intervals for the reliability in stress-strength models are discussed and empirically compared. The particular case of a bivariate normal setup is considered. The confidence intervals suggested are obtained employing approximations or asymptotic properties of maximum likelihood estimators. The coverage and the precision of these intervals are empirically checked through a simulation study. An application to real paired data is also provided.
Abstract: Recent articles have addressed the problem to construct the confidence intervals for the mean of a normal distribution where the parameter space is restricted, see for example Wang [Confidence intervals for the mean of a normal distribution with restricted parameter space. Journal of Statistical Computation and Simulation, Vol. 78, No. 9, 2008, 829–841.], we derived, in this paper, analytic expressions of the coverage probability and the expected length of confidence interval for the normal mean when the whole parameter space is bounded. We also construct the confidence interval for the normal variance with restricted parameter for the first time and its coverage probability and expected length are also mathematically derived. As a result, one can use these criteria to assess the confidence interval for the normal mean and variance when the parameter space is restricted without the back up from simulation experiments.
Abstract: The double exponential model (DEM), or Laplace
distribution, is used in various disciplines. However, there are issues
related to the construction of confidence intervals (CI), when using
the distribution.In this paper, the properties of DEM are considered
with intention of constructing CI based on simulated data. The
analysis of pivotal equations for the models here in comparisons with
pivotal equations for normal distribution are performed, and the
results obtained from simulation data are presented.