AL-Qadisiyah Journal For Administrative and Economic sciences

Abstract study examines heart disease risk factors using SCAD-penalized quantile , to capture heterogeneous covariate effects across different of disease severity while achieving effective variable selection. Unlike mean- regression models, the approach allows regression coefficients to vary quantiles of the response , providing a detailed characterization of how predictors influence mild, , and severe forms of heart disease. The framework quantile regression with the smoothly clipped deviation penalty and is through a local linear approximation algorithm. analysis is conducted on real data obtained from a publicly available heart dataset. The empirical results pronounced distributional heterogeneity in severity and highlight clear -dependent patterns. Age and ST show consistently positive and effects toward higher quantiles, indiating stronger associations among patients severe disease, while maximum rate exhibits a stable protective effect across quantiles. Other predictors, incuding resting blood pressure, serum cholesterol, exercise-induced angina, mainly at upper quantiles, suggesting their primarily for severe . Overall, the findings demonstrate that SCAD- quantile regression provides a and interpretable framework for identifying meaningful heart disease factors and uncovering heterogeneity that is not using conventional methods.

Abstract This study aims to identify the most influential factors affecting the efficiency of digital financial services in Iraq within the context of financial inclusion. It employs the Reciprocal LASSO method as a modern statistical tool for variable selection. The analysis is based on annual data from 2010 to 2020, obtained from the Financial Access Survey (FAS) published by the International Monetary Fund. Nine financial inclusion indicators were used as explanatory variables. The model results revealed that variables related to the actual use of digital technologies such as the number of online or mobile banking users, the number of debit cards, and point-of-sale (POS) terminals have the greatest impact on service efficiency. In contrast, traditional infrastructure variables like the number of bank branches showed no significant effect.These findings highlight the importance of focusing on user behavior and the adoption of digital tools rather than relying solely on quantitative expansion of banking infrastructure. The study recommends enhancing digital financial literacy, improving digital infrastructure, and providing a flexible and secure regulatory environment to boost service efficiency and achieve effective financial inclusion.

Abstract  Robust variable selection is essential in high-dimensional medical data analysis, where the presence of outliers can significantly impact model performance. This study introduces the Reciprocal Lasso, a novel regularization method that enhances robustness while preserving sparsity in regression modeling. The method incorporates an inverse penalty function that dynamically adjusts the penalization strength based on coefficient magnitudes, reducing sensitivity to extreme values. 
A comprehensive simulation study is conducted to evaluate the performance of the Reciprocal Lasso under varying levels of contamination, comparing it to the Adaptive Lasso and the S-Estimator-based Lasso. To further improve robustness, the model is integrated with Tukey’s Biweight Loss Function and MM-Estimators, which provide stronger resistance against extreme observations and improve estimation stability. The results demonstrate that the Reciprocal Lasso achieves superior variable selection accuracy, lower prediction error, and greater stability in the presence of outliers. Additionally, the method is applied to a real-world medical dataset, where it effectively identifies relevant biomarkers associated with disease progression while maintaining robustness to data anomalies. These findings suggest that the Reciprocal Lasso, combined with advanced robust estimation techniques, is a promising approach for high-dimensional modeling in medical research. Future studies could explore its application in genomic and epidemiological studies, as well as its integration with Bayesian frameworks for uncertainty quantification. 

Abstract Semiparametric Single Index Models are important and essential tools for addressing the highdimensional problem, as they play an important role in the model-building process and selecting marginal variables. In this research, some modern penal methods have been used, which assess the vector of parameters and simultaneously select the variable for the quasi-parameter single indicator models (MAVE-LASSO) and MAVE Elastic net) to improve the accuracy and predictability of the model. In order to achieve this goal, simulation experiments were conducted to demonstrate the preferences of the methods used in estimating and selecting the variable for the model. Different models, different variations, different sample sizes, and real data of factors influencing patients with blood diabetes were used for comparison and verification of the performance of these methods in practice. The methods studied will be compared by relying on two benchmarks for comparison: the average mean square error (AMSE) and the absolute average mean square error (AMAE), and the results will be obtained based on the R-code. Theoretically and simulation-wise, the MAVE-EN method for estimating and selecting important variables for a single-indicator semi-parametric model has been shown to be effective in dealing with cases of high correlation between explanatory variables in the presence of different variances.