In this paper, the Weibull extension distribution parameters are estimated under a progressive type-II censoring scheme with random removal. The parameters of the model are estimated using the maximum likelihood method, maximum product spacing, and Bayesian estimation methods. In classical estimation (maximum likelihood method and maximum product spacing), we did use the Newton–Raphson algorithm. The Bayesian estimation is done using the Metropolis–Hastings algorithm based on the square error loss function. The proposed estimation methods are compared using Monte Carlo simulations under a progressive type-II censoring scheme. An empirical study using a real data set of transformer insulation and a simulation study is performed to validate the introduced methods of inference. Based on the result of our study, it can be concluded that the Bayesian method outperforms the maximum likelihood and maximum product-spacing methods for estimating the Weibull extension parameters under a progressive type-II censoring scheme in both simulation and empirical studies.

This paper explores the efficacy of incorporating tail dependence into copula-based regression models applied to mixed health insurance data. Recognizing the limitations of traditional Generalized Linear Models (GLMs) in capturing the nuanced relationships within mixed data types, we extend the GLM framework to include bivariate and multivariate structures with gamma and negative binomial distributions. We apply a comprehensive suite of copula families—Gaussian, Clayton, Gumbel, Frank, and Student's-t to model the dependencies between variables, focusing on capturing tail dependence, a critical aspect in the context of insurance claim sizes and frequencies. Our methodology involves fitting bivariate GLMs for each pair of variables to understand pairwise dependencies and then extending the analysis to multivariate GLMs to capture the complex interplay between multiple predictors and the response variable. The analysis is performed on a rich dataset of health insurance claims, to identify the copula family that best represents the dependence structure. The results demonstrate that copulas with heavier tails, such as the Gumbel and Student's t copulas, provide superior fit and predictive performance for extreme claim amounts, outperforming those with lighter tails, such as the Gaussian and Frank copulas. The Clayton copula also shows promise in modeling lower tail dependence. Our findings suggest that tail dependence is a significant factor in accurately modeling health insurance claims data, and that the choice of copula family has a profound impact on the model's effectiveness. We conclude that copula-based regression models, with a focus on tail dependence, offer a robust alternative to conventional regression techniques, enabling actuaries and data analysts in health insurance to better understand risk and price policies more accurately. Our research contributes to the actuarial field by providing a systematic comparison of copula families in the context of health insurance data and by underscoring the importance of tail dependence in actuarial modeling

In financial markets, understanding the dynamic relationships between assets is crucial for effective portfolio management. This study highlights the importance of using the DCC-GARCH (Dynamic Conditional Correlation - Generalized Autoregressive Conditional Heteroskedasticity) model as a powerful multivariate analysis tool to capture the dynamic correlations between the S&P 500, Crude Oil Price, Natural Gas Price, and Gold Price. The DCC-GARCH model provides a flexible framework for modeling time-varying correlations, allowing investors to account for the changing relationships between assets over time. The study estimates the correlations and forecasts their evolution over the next 365 days, providing valuable insights for portfolio optimization and risk management. The results demonstrate the potential diversification benefits offered by these assets and emphasize the need for adaptive portfolio management based on the dynamic correlations. By employing the DCC-GARCH model, investors can better understand the complex interactions between assets and make more informed decisions about asset allocation, ultimately leading to improved risk-adjusted returns. This study underscores the significance of incorporating advanced multivariate techniques, such as DCC-GARCH, in financial analysis and portfolio management.

This paper investigates the efficacy of Quantile Regression, Exponentially Weighted Moving Average (EWMA), and t-GARCH models in estimating Value at Risk (VaR) for Egyptian inflation rate. Through empirical analysis and Back-testing, we demonstrate that Quantile Regression outperforms the other models in accuracy and reliability for capturing tail risks. By directly modelling the quantiles of return distributions, Quantile Regression provides a robust framework for VaR estimation, effectively addressing non-linearities and outliers in financial data. The model's ability to directly estimate quantiles allows for a nuanced understanding of extreme inflationary movements. Our findings suggest that Quantile Regression is a superior tool for risk management, offering significant advantages in precision and adaptability compared to traditional methods, which is offering valuable insights for risk managers and policymakers.

Heart diseases, are among the leading causes of mortality worldwide. Early prediction and prevention of heart disease can significantly reduce fatalities and improve patients' quality of life. In this study, we propose an advanced hybrid approach that combines multiple machine learning algorithms to predict the likelihood of heart disease in individuals. In recent times, the emergence of machine learning algorithms has shown great promise as a means to predict the risk of heart disease, including Support Vector Machines (SVM), Random Forest (RF), Decision Trees (DT), and Naïve Bayesian (NV). Our results demonstrate the effectiveness of machine learning algorithms, both individually and in combination, for heart disease diagnosis. We provide a comprehensive analysis of the strengths and weaknesses of each algorithm, as well as the ensemble models, and evaluate our approach using eight performance matrices. Our results show that the Random Forest algorithm outperforms other algorithms with an accuracy of 96%, sensitivity of 97.6%, and specificity of 94.7%. Our findings suggest, depends on the growing body of literature, the use of machine learning algorithms for heart disease diagnosis which provides valuable insights for the way for personalized and targeted interference.