Model Assisted Statistics and Applications - Volume 14, issue 1

Authors: Cordeiro, Gauss M. | Prataviera, Fábio | do Carmo S. Lima, Maria | Ortega, Edwin M.M.

Article Type: Research Article

Abstract: We study some mathematical properties of a new generator of continuous distributions with one additional shape parameter named the Marshall-Olkin extended-G (MOE-G) family. We propose a new distribution called the MOE-Flexible Weibull and obtain some of its structural properties. Further, we construct a regression model based on this distribution with two systematic components that can be applied to real lifetime data. We estimate its model parameters by maximum likelihood. We present a diagnostic analysis based on global influence and deviance residuals. Two real data sets are used to illustrate the potentiality of the proposed models.

Keywords: Censored data, flexible weibull distribution, Marshall-Olkin family, maximum likelihood estimation, regression model

DOI: 10.3233/MAS-180455

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 1-17, 2019

Authors: Adamic, Peter | Guse, Jenna | Dumais, Cedric

Article Type: Research Article

Abstract: In this paper we develop self-consistent and smoothed dependent estimators for the cause-specific failure time density in a competing risks context, employed in the presence of both left-censored and right-censored data, while allowing for masking of the failure cause. Dependence will be incorporated between the failure times and both the censoring times and the masked causes with the use of both Kernel Regression and Multivariate Multiple Regression at each iteration of the algorithm. Our approach to modeling the cause-specific failure times is intended to be the most automated and data-driven approach possible.

Keywords: Dependence modeling, multiple decrement, censoring; masking, multivariate regression

DOI: 10.3233/MAS-180456

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 19-29, 2019

Authors: Chen, Li-Pang

Article Type: Research Article

Abstract: Cross-validation (CV) and direct plug-in (DPI) are two commonly used bandwidth selection methods in nonparametric estimation. In this paper, we compare the performance of CV and DPI methods in local polynomial kernel regression through simulation study. We consider continuous response and binary response cases, with local constant, local linear and local quadratic kernel estimators, respectively. Furthermore, we investigate the first derivatives of local quadratic kernel estimators. Our results show that CV and DPI methods excel in different cases, in terms of minimizing the Mean Integrated Squared Errors (MISE); the results are also verified in empirical studies.

Keywords: Cross-validation, directly plug-in method, first-order derivative, generalized linear model, local polynomial estimator

DOI: 10.3233/MAS-180457

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 31-45, 2019

Authors: Walsh, Joe | Heazlewood, Ian Timothy | Climstein, Mike

Article Type: Research Article

Abstract: The aim of this research was to investigate statistical learning techniques to predict gender based upon psychological constructs (measuring motivations to participate in masters sports). Motivations of marathoners scales (MOMS) sports psychological data for 3928 masters athletes (2,010 males) from the World Masters Games (the largest sporting event in the world by participant numbers) was investigated using the regularized linear modelling methodologies ridge regression, the lasso and the elastic net. Comparison was made between previously published research utilizing logistic regression, discriminate function analysis, radial basis functions, multilayer perceptrons and a selection of boosted decision tree based models. It was hypothesized …that the regularized linear models would perform better than other models except the boosted decision trees, however ensembles of the regularized linear models and gradient boosted machines would result in improved accuracy over any prior models in the literature. Implementing modern regression methods with regularization provided improvements in classification accuracy based on gender compared to non-regularized linear models, however not boosted trees such as a gradient boosted machine (GBM). Models that were solely or partially based on L2 regularization (including a penalty term to reduce the sum of squares of the parameters) performed better than those than relied solely or primarily on L1 regularization (including a penalty term to reduce the sum of the absolute values of the parameters) or subset selection. This finding had implications for analysis of MOMS data in general with respect to using the 56 questions in MOMS as opposed to the underlying nine constructs for analysis in order to compensate for multicollinearity. Ensemble methods stacking ridge regression and GBMs with out-of-sample prediction further improved accuracy, giving higher accuracy scores (0.7236) than obtained in any preceding literature. This demonstrates the potential benefits from such an ensemble approach in terms of developing models with improved accuracy, as well as increasing the likelihood of developing practical applications from predictions using MOMS psychometric data. Show more

Keywords: Sport psychology, masters sport, linear modelling, ensemble modelling

DOI: 10.3233/MAS-180454

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 47-64, 2019

Authors: Latpate, R.V. | Kshirsagar, J.K.

Article Type: Research Article

Abstract: The traditional sampling methods such as simple random sampling, stratified sampling etc. cannot be used to study the rare and clustered populations. Such type of populations are frequently observed in ecological, environmental and social sciences. In such situations, often the auxiliary information is collected along with the variable of interest. Obviously, one would like to exploit this auxiliary information to the maximum extent. We consider an auxiliary variable which is highly negatively correlated with the variable of interest. An initial random sample of a fixed size is drawn from the population under study. Further, networks are formed around the units …selected in this sample that satisfy the pre specified condition with respect to the auxiliary variable. We used the procedure given by Thompson (1990) for that purpose. The variable of interest is measured corresponding to the units included in these networks. In such situation, negative adaptive cluster sampling (NACS) is of more practical interest than that of the conventional sampling designs. NACS can provide more informative sample for the investigator and more efficient estimates of the population parameters of interest. The parameters of the population are estimated by using the information on the variable of interest corresponding to the units included in the different networks. Different estimators are proposed in this article for the population total of the interest variable. The performance of these estimators is compared by using the data collected from a pilot study by using NACS method. Show more

Keywords: Auxiliary variable, negative adaptation, NACS, Hansen-Hurwitz (HH) and Horvitz-Thompson (HT) type estimators, modified ratio, regression and product type estimators

DOI: 10.3233/MAS-180452

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 65-81, 2019

Authors: Belov, Andrey | Sviridova, Olga

Article Type: Research Article

Abstract: It is shown that for a small sample size, when the maximum likelihood estimators lose its asymptotic properties, a more accurate point and interval estimation of the Poisson distribution parameter requires a direct investigation of the properties of the normalized likelihood function. Alternative point and interval estimates of the distribution parameter are obtained.

Keywords: Normalized likelihood function, Poisson type distribution, point estimate, confidence interval

DOI: 10.3233/MAS-180453

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 83-87, 2019

Authors: Bekareva, Svetlana | Meltenisova, Ekaterina | Kravchenko, Nataliya

Article Type: Research Article

Abstract: This study contains analysis of the influence of changes in the real exchange rate on the results of the monetary policies conducted by the central banks of emerging market economies targeting inflation. The analysis is based on the Taylor rule modified with the Ball rule and the Orphanides rule. The methods used included evaluating variables, such as the GDP, the inflation rate, the real interest rate, and the real exchange rate index, with the Hodrick-Prescott filter (HP-filter) and the two-step cointegration approach of Engle and Granger with structural breaks. Empiric estimates for two countries, Brazil and South Africa, based on …quarterly data during inflation targeting for 1999–2015, have been obtained. The influence of the exchange rate on the monetary policy results before targeting in the case of South Africa has been analyzed. For the latter economy, our analysis covered the period of 1960–2015. The period under consideration included the 2008 financial crisis, which accounted for the choice of the dates of breaks in the level and regime of the countries’ monetary policies. Our research showed that, although the approach undertaken by the central banks of both countries to using the interest rates for the purpose of inflation targeting was classical, both banks took the fluctuations in the exchange rates into account. A long-term relationship between the interest rate, the GDP gap, the real exchange rate, and the inflation deviation from its equilibrium value has been demonstrated for both Brazil and South Africa. The level shift took place in South Africa in the second quarter of 2008; and the shift in the level and regime in Brazil occurred in the fourth quarter of the same year. The model estimation proves that the significance of the exchange rate fluctuations has become more important since the crisis of 2008. For South Africa, we have empirically proved that after the country’s turn to inflation targeting and dismissal of the official exchange rate, the exchange rate fluctuations have still influenced the results of the national central bank’s policy. We have found the Taylor rule to be acceptable for South Africa and Brazil for a short-term period; in terms of inflation targeting, the interest rate has been positively influenced by the GDP gap and the inflation gap; the impact by the real effective exchange rate has been found to be negative. Show more

Keywords: Emerging market economy, central bank, monetary policy, inflation targeting, Taylor rule, cointegration, structural breakJEL Classification Codes: E48, E58

DOI: 10.3233/MAS-180451

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 89-102, 2019

Authors: Ermolova, M.D. | Penikas, H.I.

Article Type: Research Article

Abstract: The Central Banks discuss bank recapitalization arrangements. Markup to capital is needed because the current Basel approach is insensitive to some risks. As the Basel Committee moves from comprehensive risk modelling towards a revised, simplified, standardised approach, where one-two triggers measure risk, banking regulators increase demand for a capital add-on to meet unaccounted risks. The paper suggests the add-on estimates for the joint effect of the concentration and PD-LGD correlation risks leaving other unaccounted ones out-of-scope. The previous studies estimated add-ons for each risk separately. We show their joint impact on capital can be higher up to 5.3 times than …the sum of two taken apart. Then previous results do not provide sufficient capital for a bank. We obtain that Basel underestimates the joint risk in 1.9 times on average. We expect that our contribution will be useful at least but not last for specialised lending (e.g. real estate and project finance), where the joint effect of concentration and PD-LGD correlation risks is the most observable. Show more

Keywords: Basel, capital, add-on, concentration, PD-LGD correlation

DOI: 10.3233/MAS-180450

Citation: Model Assisted Statistics and Applications, vol. 14, no. 1, pp. 103-120, 2019