JournalTOCs

Sankhya B - Applied and Interdisciplinary Statistics
Journal Prestige (SJR): 0.1

Number of Followers: 0

Hybrid journal (It can contain Open Access articles)
ISSN (Print) 0976-8386 - ISSN (Online) 0976-8394
Published by Springer-Verlag

[2467 journals]

Joint Linear Modeling of Mixed Data and Its Application to Email Analysis
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract We present a new model in Social Networks which allows experts in this field to analyze social networks. In this paper, a joint random effect linear model for analysing longitudinal inflated [0,1]-support and inflated count response variables, where there is the possibility of non-ignorable missing values for inflated [0,1]-support response variable, has been presented. Considering the posterior distribution of unknowns given all available information. A Monte Carlo EM algorithm is used for estimating the posterior distribution of the parameters. A sensitivity of the results to the assumptions is also investigated the perturbation from missing at random to not missing at random. Influence of small perturbation of these elements on posterior displacement is also studied. Finally, for showing the applicability of the proposed model, results from analyzing Enron email dataset and student activity and profile dataset are presented. Also, a new statistical monitoring to study the longitudinal social network datasets via considering attributes which are important in various applications is provided. For this purpose, a complete definition of responsiveness rate in social networks as a [0,1]-support variable has been presented.
  PubDate: 2023-03-10
Bayesian Inference Under Ramp Stress Accelerated Life Testing Using Stan
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract In this paper, the implementations of No-U-Turn Sampler (NUTS), an extension of Hamiltonian Monte Carlo (HMC) method, via Stan software is considered for the first time under ramp stress accelerated life testing (RS-ALT). Assuming an extended Weibul (EW) distribution in the presence of adaptive type-II progressive censoring (A-II-PC) scheme, NUTS is adopted to obtain point and interval Bayesian estimation for the unknown parameters and acceleration factors when the scale parameter is related to stress through inverse power law relationship. One-sample and two-sample prediction problems are also studied under the same framework using two different approaches. To asses the performance of the suggested methods, a Monte Carlo simulation study is conducted. Finally, a real data example is provided to illustrate the application of the proposed methods in reality.
  PubDate: 2023-03-07
Modeling Long Term Return Distribution and Nonparametric Market Risk
Estimation
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract The log-return of an asset is the change in the asset price, measured in natural logarithmic scale, over a certain time period. We introduce a mathematical model for long term asset return. This model is a generalization of the well known random walk model and provides the mathematical basis for normal approximation and i.i.d. bootstrap approximation of the long-term return distribution and its quantiles. Our results yield estimators of long term value at risk (VaR) and median shortfall (MS) which are well known measures of market risk. Extensive simulations suggest that the proposed estimators outperform a number of existing estimators of VaR and MS especially over a time horizon of at least one year. Unconditional backtest by Kupiec (J. Derivat. 3, 73–84 1995) based on the annual returns of the Nifty 50 index of the national stock exchange in India, crude oil and gold prices suggests that the proposed model yields reliable estimates of the one-year Value-at-Risk and Median-Shortfall for these assets.
  PubDate: 2023-02-21
Exploring A New Class of Inequality Measures and Associated Value
Judgements: Gini and Fibonacci-Type Sequences
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract This paper explores a single-parameter generalization of the Gini inequality measure. Taking the starting point to be the Borda-type social welfare function, which is known to generate the standard Gini measure, in which incomes (in ascending order) are weighted by their inverse rank, the generalisation uses a class of non-linear functions. These are based on the so-called ‘metallic sequences’ of number theory, of which the Fibonacci sequence is the best-known. The value judgements implicit in the measures are explored in detail. Comparisons with other well-known Gini measures, along with the Atkinson measure, are made. These are examined within the context of the famous ‘leaky bucket’ thought experiment, which concerns the maximum leak that a judge is prepared to tolerate, when making an income transfer from a richer to a poorer person. Inequality aversion is thus viewed in terms of being an increasing function of the leakage that is regarded as acceptable.
  PubDate: 2023-02-21
Regression Trees and Ensemble for Multivariate Outcomes
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract Tree-based methods have become one of the most flexible, intuitive, and powerful analytic tools for exploring complex data structures. The best documented, and arguably most popular uses of tree-based methods are in biomedical research, where multivariate outcomes occur commonly (e.g. diastolic and systolic blood pressure and nerve conduction measures in studies of neuropathy). Existing tree-based methods for multivariate outcomes do not appropriately take into account the correlation that exists in such data. In this paper, we develop goodness-of-split measures for building multivariate regression trees for continuous multivariate outcomes. We propose two general approaches: minimizing within-node homogeneity and maximizing between-node separation. Within-node homogeneity is measured using the average Mahalanobis distance and the determinant of the variance-covariance matrix. Between-node separation is measured using the Mahalanobis distance, Euclidean distance and standardized Euclidean distance. To enhance prediction accuracy we extend the single multivariate regression tree to an ensemble of multivariate trees. Extensive simulations are presented to examine the properties of our goodness-of-split measures. Finally, the proposed methods are illustrated using two clinical datasets of neuropathy and pediatric cardiac surgery.
  PubDate: 2023-02-16
Generalized Double Sampling Family of Estimators for Population mean of
Sensitive Variable Harnessing Non-sensitive Auxiliary Variable and
Attribute
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract We present an enhanced double sampling generalized type estimator for the population mean of a sensitive research variable using information gathered from non-sensitive auxiliary variables and attribute, using a two-phase sampling procedure. Some special cases of the suggested family of estimators are also discussed. The expressions for bias and mean squared error of the proposed generalized estimators are derived and theoretical comparisons are made with competing estimators. Theoretical results are supported by numerical evidence generated from real-world data. A simulation analysis is also carried out to compare the efficiencies of the proposed and competing family of esimators. Both data sets show that the proposed generalized class of estimators outperforms all other estimators currently in use.
  PubDate: 2023-01-03
  DOI: 10.1007/s13571-022-00299-w
Ratio-cum-product Type Estimators for Rare and Hidden Clustered Population
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract The use of multi-auxiliary variables helps in increasing the precision of the estimators, especially when the population is rare and hidden clustered. In this article, four ratio-cum-product type estimators have been proposed using two auxiliary variables under adaptive cluster sampling (ACS) design. The expressions of the mean square error (MSE) of the proposed ratio-cum-product type estimators have been derived up to the first order of approximation and presented along with their efficiency conditions with respect to the estimators presented in this article. The efficiency of the proposed estimators over similar existing estimators have been assessed on four different populations two of which are of the daily spread of COVID-19 cases. The proposed estimators performed better than the estimators presented in this article on all four populations indicating their wide applicability and precision.
  PubDate: 2022-12-08
  DOI: 10.1007/s13571-022-00298-x
Transfer Learning in Genome-Wide Association Studies with Knockoffs
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract This paper presents and compares alternative transfer learning methods that can increase the power of conditional testing via knockoffs by leveraging prior information in external data sets collected from different populations or measuring related outcomes. The relevance of this methodology is explored in particular within the context of genome-wide association studies, where it can be helpful to address the pressing need for principled ways to suitably account for, and efficiently learn from the genetic variation associated to diverse ancestries. Finally, we apply these methods to analyze several phenotypes in the UK Biobank data set, demonstrating that transfer learning helps knockoffs discover more associations in the data collected from minority populations, potentially opening the way to the development of more accurate polygenic risk scores.
  PubDate: 2022-11-15
  DOI: 10.1007/s13571-022-00297-y
THE STATE OF ECONOMETRICS AFTER JOHN W. PRATT, ROBERT SCHLAIFER, BRIAN
SKYRMS, AND ROBERT L. BASMANN
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: SUMMARY Thirty-six years ago, introducing a distinction between factors and concomitants in regressions, John W. Pratt and Robert Schlaifer determined that the error term in a regression represents the net effect of omitted relevant regressors. As this paper demonstrates, this assumption poses a problem whenever the purpose of a model is to explain an economic phenomenon, because the estimated coefficients as well as the error will be wrong in the sense that they are not unique. But a model that is not unique cannot be a causal description of unique events in the real world. For a remedy, this paper presents a methodology based on conditions under which the error term and the coefficients on regressors included in a model do become unique, where the latter represent the sums of direct and indirect effects on the dependent variable, with omitted but relevant regressors having been chosen to define both these effects. The two effects corresponding to any particular omitted relevant regressor can be learned only by converting that regressor into an included regressor. For those cases where omitted relevant regressors are not identified, thereby preventing a meaningful distinction between direct and indirect effects, we introduce so-called coefficient drivers and a feasible method of generalized least squares, permitting a “total-effect” causal interpretation of the coefficient on each regressor included in a model.
  PubDate: 2022-11-01
  DOI: 10.1007/s13571-021-00273-y
Implications of Faithfulness in Graphical Models
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract Concentration graph models and covariance graph models are two of the widely studied classes of graphical models. They are specified through pairwise relationships between variables. Under suitable conditions, they can be used to read conditional independence relations at the level of sets of variables. It’s known that faithfulness property is filled when the graph allows identifying the whole set condition independence statements. This paper studies the implications of imposing the faithfulness assumption on either the covariance or concentration graphs. We demonstrate that if a probability distribution is faithful to its concentration graph. The corresponding covariance graph is a union of complete connected components, i.e., each connected component cannot have any marginal independence among its nodes. We also prove a dual result when the distribution is faithful to its covariance graph. The general implications of the results are far-reaching. First, the result formalizes the long-held notion in the graphical models’ community that faithfulness is a very restrictive assumption. Second, we show that estimation procedures in graphical models by low-order conditioning may lead to erroneous conclusions. Since these procedures effectively search for models in a very restrictive class of probability. distributions.
  PubDate: 2022-11-01
  DOI: 10.1007/s13571-021-00271-0
Lorenz and Polarization Orderings of the Double-Pareto Lognormal
Distribution and Other Size Distributions
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract Polarization indices such as the Foster-Wolfson index have been developed to measure the extent of clustering in a few classes with wide gaps between them in terms of income distribution. However, Zhang and Kanbur (2001) failed to empirically find clear differences between polarization and inequality indices in the measurement of intertemporal distributional changes. This paper addresses this ‘distinction' problem on the level of the respective underlying stochastic orders, the polarization order (PO) in distributions divided into two nonoverlapping classes and the Lorenz order (LO) of inequality in distributions. More specifically, this paper investigates whether a distribution F can be either more or less polarized than a distribution H in terms of the PO if F is more unequal than H in terms of the LO. Furthermore, this paper derives conditions for the LO and PO of the double-Pareto lognormal (dPLN) distribution. The derived conditions are applicable to sensitivity analyses of inequality and polarization indices with respect to distributional changes. From this application, a suggestion for appropriate two-class polarization indices is made.
  PubDate: 2022-11-01
  DOI: 10.1007/s13571-021-00264-z
Poisson Counts, Square Root Transformation and Small Area Estimation
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract The paper intends to serve two objectives. First, it revisits the celebrated Fay-Herriot model, but with homoscedastic known error variance. The motivation comes from an analysis of count data, in the present case, COVID-19 fatality for all counties in Florida. The Poisson model seems appropriate here, as is typical for rare events. An empirical Bayes (EB) approach is taken for estimation. However, unlike the conventional conjugate gamma or the log-normal prior for the Poisson mean, here we make a square root transformation of the original Poisson data, along with square root transformation of the corresponding mean. Proper back transformation is used to infer about the original Poisson means. The square root transformation makes the normal approximation of the transformed data more justifiable with added homoscedasticity. We obtain exact analytical formulas for the bias and mean squared error of the proposed EB estimators. In addition to illustrating our method with the COVID-19 example, we also evaluate performance of our procedure with simulated data as well.
  PubDate: 2022-11-01
  DOI: 10.1007/s13571-021-00269-8
COVID-19: Optimal Design of Serosurveys for Disease Burden Estimation
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract We provide a methodology by which an epidemiologist may arrive at an optimal design for a survey whose goal is to estimate the disease burden in a population. For serosurveys with a given budget of C rupees, a specified set of tests with costs, sensitivities, and specificities, we show the existence of optimal designs in four different contexts, including the well known c-optimal design. Usefulness of the results are illustrated via numerical examples. Our results are applicable to a wide range of epidemiological surveys under the assumptions that the estimate’s Fisher-information matrix satisfies a uniform positive definite criterion.
  PubDate: 2022-11-01
  DOI: 10.1007/s13571-021-00267-w
Different Coefficients for Studying Dependence
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract Through computer simulations, we research several different measures of dependence, including Pearson’s and Spearman’s correlation coefficients, the maximal correlation, the distance correlation, a function of the mutual information called the information coefficient of correlation, and the maximal information coefficient (MIC). We compare how well these coefficients fulfill the criteria of generality, power, and equitability. Furthermore, we consider how the exact type of dependence, the amount of noise and the number of observations affect their performance. According to our results, the maximal correlation is often the best choice of these measures of dependence because it can recognize both functional and non-functional types of dependence, fulfills a certain definition of equitability relatively well, and has very high statistical power when the noise grows if there are enough observations. While Pearson’s correlation does not find symmetric non-monotonic dependence, it has the highest statistical power for recognizing linear and non-linear but monotonic dependence. The MIC is very sensitive to the noise and therefore has the weakest statistical power.
  PubDate: 2022-09-01
  DOI: 10.1007/s13571-022-00295-0
Mortality Comparisons ‘At a Glance’: A Mortality Concentration Curve
and Decomposition Analysis for India
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract This paper uses the concept of the Mortality Concentration Curve (M-Curve), which plots the cumulative proportion of deaths against the corresponding cumulative proportion of the population (arranged in ascending order of age), and associated measures, to examine mortality experience in India. A feature of the M-curve is that it can be combined with an explicit value judgement (an aversion to early deaths) in order to make welfare-loss comparisons. Empirical comparisons over time, and between regions and genders, are made. Furthermore, in order to provide additional perspective, selective results for the UK and New Zealand are reported. It is also shown how the M-curve concept can be used to separate the contributions to overall mortality of changes over time (or differences between population groups) to the population age distribution and age-specific mortality rates.
  PubDate: 2022-07-28
  DOI: 10.1007/s13571-022-00293-2
A Bayesian Regression Model for the Non-standardized t Distribution with
Location, Scale and Degrees of Freedom Parameters
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract In this paper we propose Bayesian non-standardized t regression models with unknown degrees of freedom, where both location and scale parameters follow regression structures, and a Bayesian method to fit the proposed models and obtain posterior parameter inferences when the degrees of freedom are assumed to be continuous or discrete. Assuming uniform (discrete and continuous), exponential, Jeffreys and Poisson prior distributions, we develop a R-Bayesian t-regression package to obtain the posterior parameter estimates, applying a discrete and a continuous random walks in discrete and bounded real intervals, respectively. We also compare our proposal according to the usual maximum likelihood criteria, through simulations and an application to a financial dataset. In these simulations and the application, we find that our proposal has the best performance when we use the AIC, BIC and DIC criterions.
  PubDate: 2022-07-06
  DOI: 10.1007/s13571-022-00288-z
On Improving the Posterior Predictive Distribution of the Difference
Between two Independent Poisson Distribution
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract This paper addresses the exact Bayesian analysis of the difference between two independent Poisson distributions with means μ1 and μ2 respectively, known as the Skellam distribution with parameters (μ1, μ2). We develop a closed form for the posterior predictive distribution of the future distribution under the order constraint of μ1 > μ2. This kind of constraint is quite common and useful in applications specially in sports data analysis. We show that the proposed distribution estimator outperforms other types of distribution estimators in the literature. We use a simulation study with an example regarding the prediction in soccer games to show the performance of the proposed method.
  PubDate: 2022-06-07
  DOI: 10.1007/s13571-022-00284-3
Directional Measure for Analyzing the Degree of Deviance from Generalized
Marginal Mean Equality Model in Square Contingency Tables
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract When the concerned model does not fit the data, we may be interested in measuring the degree of deviance from the concerned model. This study proposes a measure for simultaneously analyzing the degree and direction of deviance from the generalized marginal mean equality model based on the ordered scores for each category. Previous study proposed a measure for analyzing both the degree and direction of deviance from the marginal mean equality model based on only the equally spaced scores. When it is appropriate to assign the ordered scores to categories, we are interested in analyzing whether the row marginal mean based on the known ordered scores is equal to the column marginal mean. It is necessary to analyze both the degree and direction of deviance from the generalized marginal mean equality model because there are two kinds of direction. We derive a confidence interval for the proposed measure using the delta method. The proposed measure is also helpful for comparing degrees of deviance from the generalized marginal mean equality model for several datasets. We show the utility of the proposed measure by applied it to real data.
  PubDate: 2022-05-12
  DOI: 10.1007/s13571-022-00283-4
Robust Moderately Clipped LASSO for Simultaneous Outlier Detection and
Variable Selection
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract Outlier detection has become an important and challenging issue in high-dimensional data analysis due to the coexistence of data contamination and high-dimensionality. Most existing widely used penalized least squares methods are sensitive to outliers due to the l2 loss. In this paper, we proposed a Robust Moderately Clipped LASSO (RMCL) estimator, that performs simultaneous outlier detection, variable selection and robust estimation. The RMCL estimator can be efficiently solved using the coordinate descent algorithm in a convex-concave procedure. Our numerical studies demonstrate that the RMCL estimator possesses superiority in both variable selection and outlier detection and thus can be advantageous in difficult prediction problems with data contamination.
  PubDate: 2022-05-11
  DOI: 10.1007/s13571-022-00279-0
Dynamic Copulas for Monotonic Dependence Change in Time Series
- Free pre-print version: Loading...
  Rate this result: What is this?
  
  Please help us test our new pre-print finding feature by giving the pre-print link a rating.
  A 5 star rating indicates the linked pre-print has the exact same content as the published article.
  
  Abstract: Abstract A particular class of dynamic bivariate copulas, monotonically increasing or decreasing, is studied for modeling dependence in a time series. As increasing or decreasing functions of time, the copula parameters are estimated via their own parameters. The method of Inference Functions for Margins (IFM), adapted from the static case, is applied for this purpose. Simulations are used to assess the detectability of an increase or a decrease in dependence over time for five copula functions. In an application to wheat prices (source: Food and Agriculture Organization), information criteria are used to select the best copula function, and the dynamic copulas are shown to represent an improvement over static copulas for several of the time series.
  PubDate: 2022-05-11
  DOI: 10.1007/s13571-022-00281-6