Search

Given the need to make efficient use of the resources allocated to experimental research in sugarcane crops, this research was proposed with the aim of determining the optimal size of the experimental plot in sugarcane crops in the Brunca region of Costa Rica. During the 2018-2019 harvest, an experimental uniformity trial was established, and five methods were compared: maximum curvature method, maximum curvature of the coefficient of variation, linear regression with constant, quadratic regression with constant and maximum distance method. The results indicate that the most efficient estimators were obtained with models that consider all sizes and forms of the uniformity trial (n=63). Segmented regression and linear regression with constant models produced the best estimators of the optimal size of the experimental plot: 72.16 and 93.22 m², respectively. With the other three methods, considerable and inconsistent differences in the sizes of the experimental plot were obtained. With the methods of maximum curvature and maximum curvature of the coefficient of variation, the results were so small: 14.01 and 12.5 m², respectively, that they are inadequate to carry out research in sugarcane; on the contrary, with the method of maximum distance, the size obtained was 157.48 m², statistically and economically inefficient. Therefore, the linear regression with constant and quadratic regression with constant models are appropriate for determining the experimental plot size in sugarcane. It was concluded that the recommended size to be used in the area is 72 m². This research was completed in December 2021. [ABSTRACT FROM AUTHOR]

The Ordinary Least Square (OLS) estimator remains Best Linear Unbiased Estimator (BLUE) when all the assumptions surrounding it stay intact, but at an iota of violation of the assumptions, it becomes inefficient and unstable. Some causes of the violation are the multicollinearity and the presence of extreme values (outliers). Recently robust Kibria -Lukman based on M estimator was proposed by Majid et al. (2022) but when there are outlying cases in the y-direction. Since, outliers in the x-direction may be inevitable in the data set, therefore it becomes imperative to examine the performance of the robust-M Kibria-Lukman (KL-M) estimator as alternative to already proposed robust estimators that can handle these problems when there are outliers in the x-direction. Through the Monte Carlo experiment, theoretical results under some conditions and factors, including application to real-life data, the new estimator outperformed other estimators considered in this study in the presence of multicollinearity and extreme values in the x-direction. As the error variances (σ2), level of multicollinearity (rho) and percentage (px), and magnitude (mx) of outliers increase, the Mean Square Error (MSE) of the estimators' considered increase. Meanwhile the MSEs of the estimators decrease as the sample size (n) increases. When rho>0, mx>0, the (px) increases, and sample size (n) increases KL-M along sides, ordinary Kibria-Lukman (KL) estimator outperformed other estimators as the two anomalies occur simultaneously. The KL-M performed better, especially when the sample size was n=100. Conclusively, at the different biasing parameters of the estimators, KL-M performed better than other estimators considered in the study. In the same vein, real-life data was adopted to affirm the claim. [ABSTRACT FROM AUTHOR]

Causal decomposition analysis is among the rapidly growing number of tools for identifying factors ("mediators") that contribute to disparities in outcomes between social groups. An example of such mediators is college completion, which explains later health disparities between Black women and White men. The goal is to quantify how much a disparity would be reduced (or remain) if we hypothetically intervened to set the mediator distribution equal across social groups. Despite increasing interest in estimating disparity reduction and the disparity that remains, various estimation procedures are not straightforward, and researchers have scant guidance for choosing an optimal method. In this article, the authors evaluate the performance in terms of bias, variance, and coverage of three approaches that use different modeling strategies: (1) regression-based methods that impose restrictive modeling assumptions (e.g., linearity) and (2) weighting-based and (3) imputation-based methods that rely on the observed distribution of variables. The authors find a trade-off between the modeling assumptions required in the method and its performance. In terms of performance, regression-based methods operate best as long as the restrictive assumption of linearity is met. Methods relying on mediator models without imposing any modeling assumptions are sensitive to the ratio of the group-mediator association to the mediator-outcome association. These results highlight the importance of selecting an appropriate estimation procedure considering the data at hand. [ABSTRACT FROM AUTHOR]

Copyright of GeSec: Revista de Gestao e Secretariado is the property of Sindicato das Secretarias e Secretarios do Estado de Sao Paulo (SINSESP) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Entropy estimation is a fundamental problem in information theory that has applications in various fields, including physics, biology, and computer science. Estimating the entropy of discrete sequences can be challenging due to limited data and the lack of unbiased estimators. Most existing entropy estimators are designed for sequences of independent events and their performances vary depending on the system being studied and the available data size. In this work, we compare different entropy estimators and their performance when applied to Markovian sequences. Specifically, we analyze both binary Markovian sequences and Markovian systems in the undersampled regime. We calculate the bias, standard deviation, and mean squared error for some of the most widely employed estimators. We discuss the limitations of entropy estimation as a function of the transition probabilities of the Markov processes and the sample size. Overall, this paper provides a comprehensive comparison of entropy estimators and their performance in estimating entropy for systems with memory, which can be useful for researchers and practitioners in various fields. [ABSTRACT FROM AUTHOR]

In this paper, we propose a methodology to evaluate the effects on the value-added estimators when there is a measurement error in variables. These errors lead to biased estimators; in most cases, these estimators have a large variance. The proposed methodology is illustrated under two scenarios: First, simulation studies for discussing the effects on value-added estimates for some values of reliability. Second, an application for real data using a Colombian database that contains the scores of two standardized tests: Saber 11 and Saber Pro. In the first test, the students are evaluated in the 11th grade of high school. In the second test, the students are evaluated in the last two years of the undergraduate program. We also implement the estimation method taking into account measurement error and adapt the Bootstrap procedure to Hierarchical Linear Mixed Model with error in variables. The results show that measurement error affects the higher-education value-added and the estimator variance. This implies universities can be incorrectly classified. Therefore, we may assert that any given university is contributing to students' progress when, in reality, this is not so. [ABSTRACT FROM AUTHOR]

Risk measurements are clearly central to risk management, in particular for banks, (re)insurance companies, and investment funds. The question of the appropriateness of risk measures for evaluating the risk of financial institutions has been heavily debated, especially after the financial crisis of 2008/2009. Another concern for financial institutions is the pro-cyclicality of risk measurements. In this paper, we extend existing work on the pro-cyclicality of the Value-at-Risk to its main competitors, Expected Shortfall, and Expectile: We compare the pro-cyclicality of historical quantile-based risk estimation, taking into account the market state. To characterise the latter, we propose various estimators of the realised volatility. Considering the family of augmented GARCH(p , q) processes (containing well-known GARCH models and iid models, as special cases), we prove that the strength of pro-cyclicality depends on the three factors: the choice of risk measure and its estimators, the realised volatility estimator and the model considered, but, no matter the choices, the pro-cyclicality is always present. We complement this theoretical analysis by performing simulation studies in the iid case and developing a case study on real data. [ABSTRACT FROM AUTHOR]

We consider a stationary linear model with zero mean. The autoregression parameters, as well as the distribution function (d.f.) of innovations, are unknown. We test symmetry of innovations with respect to zero in two situations. In the first case the observations are a sample from a stationary solution of . We estimate parameters and find residuals. Based on them we construct a kind of empirical d.f. and the omega-square type test statistic. Its asymptotic d.f. under the hypothesis and the local alternatives are found. In the second situation the observations are subject to gross errors (outliers). For testing the symmetry of innovations we again construct the Pearson's type statistic and find its asymptotic d.f. under the hypothesis and the local alternatives. We establish the asymptotic robustness of Pearson's test as well. [ABSTRACT FROM AUTHOR]

The multinomial distribution is often used in modeling categorical data because it describes the probability of a random observation being assigned to one of several mutually exclusive categories. Given a finite or numerable multinomial model M | n , p whose decision is indexed by a parameter θ and having a cost c θ , p depending on θ and on p , we show that, under general conditions, the probability of taking the least cost decision tends to 1 when n tends to ∞, i.e., we showed that the cost decision is consistent, representing a Statistical Decision Theory approach to the concept of consistency, which is not much considered in the literature. Thus, under these conditions, we have consistency in the decision making. The key result is that the estimator p ˜ n with components p ˜ n , i = n i n , i = 1 , ⋯ , where n i is the number of times we obtain the ith result when we have a sample of size n, is a consistent estimator of p. This result holds both for finite and numerable models. By this result, we were able to incorporate a more general form for consistency for the cost function of a multinomial model. [ABSTRACT FROM AUTHOR]

Most reported treatment effects in medical research studies are ambiguously defined, which can lead to misinterpretation of study results. This is because most authors do not attempt to describe what the treatment effect represents, and instead require readers to deduce this based on the reported statistical methods. However, this approach is challenging, because many methods provide counterintuitive results. For example, some methods include data from all patients, yet the resulting treatment effect applies only to a subset of patients, whereas other methods will exclude certain patients while results will apply to everyone. Additionally, some analyses provide estimates pertaining to hypothetical settings in which patients never die or discontinue treatment. Herein we introduce estimands as a solution to the aforementioned problem. An estimand is a clear description of what the treatment effect represents, thus saving readers the necessity of trying to infer this from study methods and potentially getting it wrong. We provide examples of how estimands can remove ambiguity from reported treatment effects and describe their current use in practice. The crux of our argument is that readers should not have to infer what investigators are estimating; they should be told explicitly. [ABSTRACT FROM AUTHOR]

We propose a novel family of estimators for the population mean under non-response and obtain the MSE equation of the suggested estimator for each situation in theory. These theoretical conditions are applied to three popular data sets in literature and we see that the suggested estimators are more efficient than the traditional estimators, such as ratio, regression estimators, in Case 1; whereas, in Case 2, the suggested estimators are also more efficient than the Unal-Kadilar exponential estimators that are more efficient than the traditional estimators for the same data sets. [ABSTRACT FROM AUTHOR]

Entropy estimation is a fundamental problem in information theory that has applications in various fields, including physics, biology, and computer science. Estimating the entropy of discrete sequences can be challenging due to limited data and the lack of unbiased estimators. Most existing entropy estimators are designed for sequences of independent events and their performances vary depending on the system being studied and the available data size. In this work, we compare different entropy estimators and their performance when applied to Markovian sequences. Specifically, we analyze both binary Markovian sequences and Markovian systems in the undersampled regime. We calculate the bias, standard deviation, and mean squared error for some of the most widely employed estimators. We discuss the limitations of entropy estimation as a function of the transition probabilities of the Markov processes and the sample size. Overall, this paper provides a comprehensive comparison of entropy estimators and their performance in estimating entropy for systems with memory, which can be useful for researchers and practitioners in various fields.

In this work, a nonlinear estimator-based robust controller is designed for the position and yaw control of a quadrotor with uncertainty estimation. This controller ensures the tracking of desired references in the presence of parameters variation and external disturbances, making use of high-order sliding mode (HOSM) estimators to estimate these perturbations that can be canceled by the control, thus improving the dynamic behavior of the controlled system. Its performance is evaluated making use of a Simcenter Amesim quadrotor based on physical models generated from experimental data in a co-simulation framework with Matlab–Simulink used to implement the designed controller with FPGA implementation. A challenging and generic maneuver with time-varying wind disturbances and uncertainty model parameters is considered. [ABSTRACT FROM AUTHOR]

In the presence of multicollinearity and outliers, the ordinary least squares estimator remains inconsistent and unreliable. Several estimators have been proposed that can co-handle the problems of multicollinearity and outliers simultaneously. However, there is still a need to explore some other robust methods when the two anomalies appear in the linear regression model and recommend it to end users of statistics. Therefore, this study proposed Robust-M New Two Parameter (RNTP) and examined its performance over some already existing ones in the presence of multicollinearity and outliers in the x-direction. The theoretical expression under some conditions was established to showcase the new estimator's superiority. A simulation study was carried out alongside some factors to show that the RNTP is better than all other estimators considered in the study. The simulation study results revealed that RNTP outperformed other estimators in the study using the minimum MSE as the criterion. Likewise, real-life data was applied to affirm this claim.

The multinomial distribution is often used in modeling categorical data because it describes the probability of a random observation being assigned to one of several mutually exclusive categories. Given a finite or numerable multinomial model M|n,p whose decision is indexed by a parameter θ and having a cost cθ,p depending on θ and on p, we show that, under general conditions, the probability of taking the least cost decision tends to 1 when n tends to ∞, i.e., we showed that the cost decision is consistent, representing a Statistical Decision Theory approach to the concept of consistency, which is not much considered in the literature. Thus, under these conditions, we have consistency in the decision making. The key result is that the estimator p˜n with components p˜n,i=nin,i=1,⋯, where ni is the number of times we obtain the ith result when we have a sample of size n, is a consistent estimator of p. This result holds both for finite and numerable models. By this result, we were able to incorporate a more general form for consistency for the cost function of a multinomial model.

Copyright of Journal of Administration & Economics is the property of Republic of Iraq Ministry of Higher Education & Scientific Research (MOHESR) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

We display the power Topp-Leone (PTL) distribution with two parameters. The following major features of the PTL distribution are investigated: quantile measurements, certain moment's measures, residual life function, and entropy measure. Maximum likelihood, least squares, Cramer von Mises, and weighted least squares approaches are used to estimate the PTL parameters. A numerical illustration is prepared to compare the behavior of the achieved estimates. Data analysis is provided to scrutinize the flexibility of the PTL model matched with Topp-Leone distribution. [ABSTRACT FROM AUTHOR]

Phenotypic plasticity index (PI), the slope of reaction norm (K) and relative distances plasticity index (RDPI), the most commonly used estimators, have occasionally been found to generate different plasticity rankings between groups (species, populations, cultivars or genotypes). However, no effort has been made to determine how frequent this incongruence is, and the factors that influence the occurrence of the incongruence. To address these problems, we first proposed a conceptual framework and then tested the framework (its predictions) by reanalyzing 1248 sets of published data. Our framework reveals inherent conflicts between K and PI or RDPI when comparing plasticity between two groups, and the frequency of these conflicts increases with increasing inter-group initial trait difference and/or K values of the groups compared. More importantly, the estimators also affect the magnitude of the inter-group plasticity differences even when they do not change groups' plasticity rankings. The above-mentioned effects of plasticity estimators were confirmed by our empirical test using data from the literature, and the conflicts occur in 203 (16%) of the 1248 comparisons between K and indices, indicating that a considerable proportion of the comparative conclusions on plasticity in literature are estimator-dependent. The frequency of the conflicts is influenced by phylogenetic relatedness of the groups compared, being lower when comparing within relative to between species, but not by specific types of environments, traits and species. Our study indicates that care is needed to select estimator when comparing groups' plasticity, and that the conclusions in relevant literature should be treated with great caution. [ABSTRACT FROM AUTHOR]

The life risk factor for the cancer disease is becoming more and more in the present living life style. Many researchers focused on skin cancer using machine learning approaches, but when working with multi classification data, it is observed that ANN gives better results, if the input data is keeps on adding in the real time scenario. This proposed system focuses on the classification of the skin cancer types by designing the neural network based on the customized activation function. Many parameters are involved in designing a suitable classifier using NN like learning rate, optimizer, activator, and other normalization layers. This paper majorly focuses on the activation function and number of neurons associated with the layer because these two parameters play a vital role in the entire accuracy of the model. Among the existing activators, the combination of tanh and relu has given high value base on them; a new activation function is designed. [ABSTRACT FROM AUTHOR]

In this work, a nonlinear estimator-based robust controller is designed for the position and yaw control of a quadrotor with uncertainty estimation. This controller ensures the tracking of desired references in the presence of parameters variation and external disturbances, making use of high-order sliding mode (HOSM) estimators to estimate these perturbations that can be canceled by the control, thus improving the dynamic behavior of the controlled system. Its performance is evaluated making use of a Simcenter Amesim quadrotor based on physical models generated from experimental data in a co-simulation framework with Matlab–Simulink used to implement the designed controller with FPGA implementation. A challenging and generic maneuver with time-varying wind disturbances and uncertainty model parameters is considered.

The present article aims to shows that obtaining the best estimator does not depend on estimators' terms and functions. To prove this situation, it is shown that there is no difference between the existing estimators and the proposed estimator by proposing new sophisticated estimators. Besides, theoretical, graphical, empirical and simulation studies are conducted to confirm the efficiency of these proposed estimators with the help of the method used by Wolter (2007). Also, the MSE values are evaluated by using the power coefficient of the interested estimators. In data sets with the known correlation coefficient, it is possible to determine which power coefficient estimator is better without computing the MSE value. [ABSTRACT FROM AUTHOR]

Multivariate exponentially weighted moving average (MEWMA) control chart with five different estimators as population covariance matrix is rarely applied to monitor small fluctuations in the statistical process control. In this article, mathematical models of the five estimators (S1, S2, S3, S4, S5) are established, with which the relevant MEWMA control charts are obtained, respectively. Thereafter, the process monitoring performance of the five control charts is simulated. And the simulation results show that the S4 estimator-based MEWMA control chart is of the best performance both in step offset failure mode and ramp offset failure mode. Since the inline process monitoring of photovoltaic manufacturing is intended to be a problem of multivariate statistics process analysis, the feasibility and effectiveness of the proposed model are elaborated in the case study during the cell testing and sorting process control for the fabrication of multicrystalline silicon solar cells. [ABSTRACT FROM AUTHOR]

Old metallic bridges are prone to high levels of structural degradation due to their long service life, therefore, it is essential to study strengthening techniques. Implementation of injection bolts to replace riveted joints exposed to fatigue damages has been studied. The mechanical performance of injection bolts has been studied mainly based on quasi‐static or creep tests. The objective of this paper is comparison of fatigue resistance between bolted joints with preloaded injection bolts opposing bolted joints with preloaded high strength bolts. Specimens were defined aiming to reproduce as close as possible a strengthening scenario. Single and double lap joints were tested. The fatigue phenomenon has a stochastic nature and therefore robust models are necessary to estimate parameters of distribution functions. Therefore, different estimation methodologies were implemented to determine the parameters of Weibull distribution. The main goal was to define reliable S‐N design curves. Results showed that the use of preloaded injection bolts contribute to increase the fatigue life and to reduce the scatter, mainly for double lap joints. It is shown that S‐N curves of EN1993‐1‐9 do not represent the fatigue strength of structural details with old metallic materials. [ABSTRACT FROM AUTHOR]

We use simple examples to show how the bias and standard error of an estimator depend in part on the type of estimator chosen from among parametric, nonparametric, and semiparametric candidates. We estimated the cumulative distribution function in the presence of missing data with and without an auxiliary variable. Simulation results mirrored theoretical expectations about the bias and precision of candidate estimators. Specifically, parametric maximum likelihood estimators performed best but must be "omnisciently" correctly specified. An augmented inverse probability–weighted (IPW) semiparametric estimator performed best among candidate estimators that were not omnisciently correct. In one setting, the augmented IPW estimator reduced the standard error by nearly 30%, compared with a standard Horvitz-Thompson IPW estimator; such a standard error reduction is equivalent to doubling the sample size. These results highlight the gains and losses that can be incurred when model assumptions are made in any analysis. [ABSTRACT FROM AUTHOR]

Abstract In this paper, estimators of the Nakagami-lognormal (NL) distribution based on the method of log-moments have been derived and thoroughly analyzed. Unlike maximum likelihood (ML) estimators, the log-moment estimators of the NL distribution are obtained using straightforward equations with a unique solution. Also, their performance has been evaluated using the sample mean, confidence regions and normalized mean square error (NMSE). The NL distribution has been extensively used to model composite small-scale fading and shadowing in wireless communication channels. This distribution is of interest in scenarios where the small-scale fading and the shadowing processes cannot be easily separated such as the vehicular environment.

The evaluation of covariance matrix estimators is very important for portfolio analysis and risk management. The Monte Carlo synthetic sample based performance evaluation method proposed by this article can avoid the main shortcomings of statistical and economic methods which are widely used in the existing literature. The proposed method does not need the true covariance and does not need to introduce the performance of the out-of-sample portfolios. It is an intuitive, effective and robust measure for both simulation and empirical analysis. [ABSTRACT FROM AUTHOR]

Researchers face many choices when conducting large-scale multisite individually randomized control trials. One of the most common quantities of interest in multisite RCTs is the overall average effect. Even this quantity is non-trivial to define and estimate. The researcher can target the average effect across individuals or sites. Furthermore, the researcher can target the effect for the experimental sample or a larger population. If treatment effects vary across sites, these estimands can differ. Once an estimand is selected, an estimator must be chosen. Standard estimators, such as fixed-effects regression, can be biased. We describe 15 estimators, consider which estimands they are appropriate for, and discuss their properties in the face of cross-site effect heterogeneity. Using data from 12 large multisite RCTs, we estimate the effect (and standard error) using each estimator and compare the results. We assess the extent that these decisions matter in practice and provide guidance for applied researchers. [ABSTRACT FROM AUTHOR]

The study covered four (4) selected financial deepening estimators as it affects the economy as a whole. To carry out the study, data were sourced through Statistical Bulletin 2018 and World Bank Development Indicators (2018) from 2003 to 2018. Group of low-income countries have continued to find the need for adequate financial deepening in a bid to enhance their economic growth. Though they are operating a modern type of economy, yet they are mostly mono-economy, relying more on either oil or agricultural products and heavily depending on oil. Hence, Nigeria is a small open economy. In analyzing the data obtained, linear regression analysis was used through SPSS 22.0. The study formulated four (4) hypotheses and the findings showed that the ratio of money supply, credit ratio in private sector, savings ratio and investment have impact significantly on Gross Domestic Product in Nigeria because the p-value t-statistics are 0.0481, 0.027, 0.046 and.000 respectively are all less than 5% significant level. The study, therefore, identified the nexus by concluding that financial deepening has significant impact on Nigeria's economic growth. This study recommends that government policies should be geared towards increased money supply and well-organized capital market that can improve general economic efficacy. [ABSTRACT FROM AUTHOR]

The paper addresses the problem of constructing lower and upper estimators for univariate functions. This problem is of crucial importance in global optimization, where such bounds are used to reduce the search area. We propose to use piecewise linear estimators for bounding univariate functions and show how such estimators can be derived from the function's algebraic expression. The basic properties of such estimators are formulated and proved. We implemented the algorithms for the automated construction of lower and upper piecewise linear estimators and experimentally compared the proposed approach with the first-order interval bounds, Pijavskij method, and slope arithmetic. Numerical examples demonstrate that the piecewise linear estimators are more accurate with respect to the mentioned approaches. We also show that global optimization algorithms can significantly benefit from using piecewise linear estimators. Another advantage of the proposed approach is that the objective function does not have to be differentiable. This feature can favorably distinguish this method from other methods where the first and second derivatives are used. [ABSTRACT FROM AUTHOR]

In the present study, generalized classes of estimators for estimating population mean, ratio and product of two population means using rank of auxiliary character in presence of non-response are proposed. The bias and mean square error of proposed classes of estimators are obtained and their performances examined. Specific conditions under which the members of proposed classes of estimators attain minimum mean square error are obtained. Comparative study of the proposed classes of estimators with the relevant estimators is carried out. An empirical study is given to justify the efficiency of the proposed classes of estimators. [ABSTRACT FROM AUTHOR]

Completeness of BMS (Butterfly Monitoring Scheme) butterfly censuses in Spain has been calculated from a total of 86 transects distributed in 21 Spanish provinces. The number of species for each transect was estimated using both parametric (rational function) and non-parametric estimators (Chao 1 and 2, Jackknife 1 and 2, Bootstrap, ICE, ACE). Results show that 75% of the transects have a completeness greater than 70%, and there are no significant differences between estimators. Correlation of completeness with other variables was also studied, the most related being number of visits, number of individuals and number of species of each transect. Finally, geographical distribution of richness and completeness was analysed and only richness resulted to be correlated with longitude and latitude, which discards the idea of a geographic bias in completeness. [ABSTRACT FROM AUTHOR]

In this article, we have highlighted limitations of existing structures of X ¯ control chart for unknown parameters by considering various circumstances of a process. The circumstances include availability of limited samples for estimating control limits, probability distribution is unknown and collected data are highly skewed. To tackle with the limitations, we have proposed generalized skewness correction structure of X ¯ chart. For proposing the required structure, we have developed skewness correction based dispersion estimators and corrected control limits multipliers to replace with known probability distribution based dispersion estimator and control limits multipliers. The proposed generalized skewness structure is dependent on the amount of skewness of gathered data from an ongoing process instead of restricted assumptions. Results illustrate that actual false alarm rate of proposed structure remains close to true false alarm rate as compared to existing structures when assumptions are violated. Besides, a real-life example from petrochemical process is presented for explaining the implementation procedure of proposed structure. [ABSTRACT FROM AUTHOR]

The quantized neural networks (QNNs) can be useful for neural network acceleration and compression, but during the training process they pose a challenge: how to propagate the gradient of loss function through the graph flow with a derivative of 0 almost everywhere. In response to this non-differentiable situation, we propose a novel Asymptotic-Quantized Estimator (AQE) to estimate the gradient. In particular, during back-propagation, the graph that relates inputs to output remains smoothness and differentiability. At the end of training, the weights and activations have been quantized to low-precision because of the asymptotic behaviour of AQE. Meanwhile, we propose a M-bit Inputs and N-bit Weights Network (MINW-Net) trained by AQE, a quantized neural network with 1–3 bits weights and activations. In the inference phase, we can use XNOR or SHIFT operations instead of convolution operations to accelerate the MINW-Net. Our experiments on CIFAR datasets demonstrate that our AQE is well defined, and the QNNs with AQE perform better than that with Straight-Through Estimator (STE). For example, in the case of the same ConvNet that has 1-bit weights and activations, our MINW-Net with AQE can achieve a prediction accuracy 1.5% higher than the Binarized Neural Network (BNN) with STE. The MINW-Net, which is trained from scratch by AQE, can achieve comparable classification accuracy as 32-bit counterparts on CIFAR test sets. Extensive experimental results on ImageNet dataset show great superiority of the proposed AQE and our MINW-Net achieves comparable results with other state-of-the-art QNNs. [ABSTRACT FROM AUTHOR]

Studies included in meta-analysis can produce results that depart from the true population parameter of interest due to systematic and/or random errors. Synthesis of these results in meta-analysis aims to generate an estimate closer to the true population parameter by minimizing these errors across studies. The inverse variance heterogeneity (IVhet), quality effects and random effects models of meta-analysis all attempt to do this, but there remains controversy around the estimator that best achieves this goal of reducing error. In an attempt to answer this question, a simulation study was conducted to compare estimator performance. Five thousand iterations at 10 different levels of heterogeneity were run, with each iteration generating one meta-analysis. The results demonstrate that the IVhet and quality effects estimators, though biased, have the lowest mean squared error. These estimators also achieved a coverage probability at or above the nominal level (95%), whereas the coverage probability under the random effects estimator significantly declined (<80%) as heterogeneity increased despite a similar confidence interval width. Based on our findings, we would recommend the use of the IVhet and quality effects models and a discontinuation of traditional random effects models currently in use for meta-analysis. [ABSTRACT FROM AUTHOR]