Article
Statistics & Probability
M. I. Alheety, Kristofer Mansson, B. M. Golam Kibria
Summary: In logistic regression model, multicollinearity can cause the variance of the maximum likelihood estimator to be inflated and unstable. A new stochastic restricted biased estimator is proposed to address this issue and its performance is compared with existing estimators through statistical properties and scalar mean squared criterion.
JOURNAL OF APPLIED STATISTICS
(2021)
Article
Computer Science, Interdisciplinary Applications
Muhammad Qasim, Muhammad Nauman Akram, Muhammad Amin, Kristofer Mansson
Summary: This article introduces a restricted gamma ridge regression estimator (RGRRE) to address multicollinearity issues in estimating the parameter beta in the gamma regression model. The properties and superiority of the new estimator are theoretically analyzed, along with suggested methods for finding the optimal shrinkage parameter value. Monte Carlo simulation and empirical application demonstrate the benefits of RGRRE over existing estimators.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2022)
Article
Computer Science, Interdisciplinary Applications
Muhammad Nauman Akram, Muhammad Amin, Adewale F. Lukman, Saima Afzal
Summary: This article introduces the inverse Gaussian regression model and proposes a principal component ridge estimator for the model in the presence of multicollinearity. Through Monte Carlo simulation study and analysis of a real dataset, it is demonstrated that the proposed estimator outperforms the classical MLE and other biased estimation methods.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2022)
Article
Computer Science, Software Engineering
Abdul Majid, Shakeel Ahmad, Muhammad Aslam, Muhammad Kashif
Summary: This article proposes a robust version of the Kibria-Lukman estimator (KLE) to address the problems of multicollinearity and outliers in regression models. The performance of the proposed method is evaluated through Monte Carlo simulation and real-life data, demonstrating its superiority.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2023)
Article
Computer Science, Information Systems
Mohamad Abou Houran, Mohamed H. Essai Ali, Adel B. Abdel-Raman, Eman A. Badry, Alaaeldien Hassan, Hany A. Atallah
Summary: This paper suggests improving the performance of Deep Learning Long Short-Term Memory (DLLSTM) structures by using robust loss functions and creating new classification layers. The effectiveness of the suggested DLLSTM classifier was examined using three loss functions (Crossentropy, MAE, SSE) for two different applications. The results show that the suggested classifier with SSE loss function outperforms others and the suggested activation functions are more accurate than the tanh function.
Article
Statistics & Probability
Nileshkumar H. Jadhav
Summary: Poisson regression is a commonly used technique for modeling count data, but maximum likelihood estimation can be unstable. A new estimator is proposed in this study to address multicollinearity issues, showing superior performance in terms of mean squared error.
JOURNAL OF APPLIED STATISTICS
(2022)
Article
Statistics & Probability
Muhammad Naumanm Akram, Muhammad Amin, Muhammad Amanullah
Summary: This paper proposes two new parameter estimators for the inverse Gaussian regression model (IGRM) to improve the efficiency of estimates. Through Monte Carlo simulation and real examples, it has been shown that these new estimators outperform other methods.
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
(2022)
Article
Statistics & Probability
Jibo Wu
Summary: This paper presents the generalized restricted difference-based almost unbiased ridge estimator in partially linear model, considering the possibility of restriction of the regression parameters to a subspace, and compares the proposed estimators using the quadratic bias and scalar mean squared error criteria. Finally, a simulation study is conducted to explain the performances of the estimators.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2022)
Article
Computer Science, Interdisciplinary Applications
Muhammad Nauman Akram, B. M. Golam Kibria, Mohamed R. Abonazel, Nimra Afzal
Summary: This study proposes some new shrinkage parameters for the gamma regression model, and demonstrates their superiority in small dispersion levels through empirical comparisons.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2022)
Article
Mathematics
Kadri Ulas Akay, Esra Ertan
Summary: In this study, a new improved Liu-type estimator is proposed to address the issue of unstable parameter estimates in the Poisson Regression Model (PRM). Through Monte Carlo simulation studies and real data analysis, the proposed estimator is shown to outperform other biased estimators in terms of performance.
HACETTEPE JOURNAL OF MATHEMATICS AND STATISTICS
(2022)
Article
Statistics & Probability
Jibo Wu, Yasin Asar
Summary: This article discusses parameter estimation in a linear measurement error model with ill-conditioned data and proposes a new two-parameter estimator to address the problem of multicollinearity. The asymptotic properties of the new estimator are considered using the mean squared error matrix. Finally, a Monte Carlo simulation is conducted to demonstrate the performance of the estimators based on simulated mean squared error criteria.
Article
Computer Science, Interdisciplinary Applications
Abdul Majid, Muhammad Amin, Muhammad Nauman Akram
Summary: The recent development of the Bell Liu regression model aims to address the issue of multicollinearity in the Bell regression model and introduces new Liu parameters. Through Monte Carlo simulation studies and real-world applications, the superiority of the proposed method is demonstrated.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2022)
Article
Statistics & Probability
Junmei Zhou, Liqin Li
Summary: The two-parameter gamma distribution, an important probability distribution with moderate skewness, is extensively used in statistics. However, the maximum likelihood estimators (MLEs) of its parameters lack closed forms, posing challenges for practical implementation. Additionally, the MLE of its shape parameter is characterized by low estimation efficiency due to significant bias. Consequently, alternative estimators have been explored in existing literature. This article proposes an easily computable MLE technique, with the modified MLE of the shape parameter demonstrating high efficiency and outperforming most existing estimators.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2023)
Article
Engineering, Aerospace
Seokkwon Kim, Sung-Wan Kim, Keunsu Ma
Summary: This article proposes a method for estimating the bit error probability of received data using a weighted least squares estimator. The characteristics of bias and variance are analyzed, and optimal weights are derived to minimize the variance. Results show that the proposed estimator has lower computational complexity and slightly larger or even less mean squared error compared to existing methods.
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS
(2021)
Article
Computer Science, Software Engineering
Iqra Babar, Sohail Chand
Summary: New weighted ridge and Liu estimators are proposed in this article to overcome the problem of multicollinearity, and they perform better than existing methods in terms of performance. The effectiveness and practicality of these methods are validated through Monte-Carlo simulations and real-life applications.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2022)
Article
Statistics & Probability
Muhammad Amin, Muhammad Nauman Akram, Abdul Majid
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
(2023)
Article
Statistics & Probability
Muhammad Nauman Akram, Muhammad Amin, Muhammad Faisal
Summary: This study proposes a generalized class of biased estimators for addressing the multicollinearity problem in gamma regression models. Through simulation and empirical applications, it is found that the generalized gamma ridge regression estimator outperforms MLE and generalized gamma Liu estimator.
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
(2023)
Article
Statistics & Probability
Abdul Majid, Muhammad Amin, Muhammad Aslam, Shakeel Ahmad
Summary: The ridge regression estimator is widely used in multiple linear regression models, but it may have adverse effects when outliers are present. To address this, robust ridge estimators based on M-estimator are proposed, which are less sensitive to outliers. This study introduces robust ridge estimators for the ridge parameter and evaluates their performance through simulations and a real application. The results show that the proposed robust ridge estimators outperform the traditional methods in terms of mean squared error.
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
(2023)
Article
Mathematics
Adewale F. F. Lukman, B. M. Golam Kibria, Cosmas K. K. Nziku, Muhammad Amin, Emmanuel T. T. Adewuyi, Rasha Farghali
Summary: The Logistic Kibria-Lukman estimator (LKLE) is proposed to handle multicollinearity in the logistic regression model. The superiority condition of this new estimator over other methods is theoretically established and validated using real-life data and simulation. The performance of the estimators is contingent on the adopted shrinkage parameter estimators.
Article
Statistics & Probability
Muhammad Nauman Akram, Nimra Afzal, Muhammad Amin, Asia Batool
Summary: The ZINB modified ridge type (ZINBMRT) estimator is proposed to address the issue of multicollinearity in the ZINB regression model. New approaches to estimate the shrinkage parameters for the ZINBMRT estimator are also proposed. Simulation and example studies show that the proposed ZINBMRT estimator outperforms other competitive estimators.
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
(2023)
Article
Computer Science, Interdisciplinary Applications
Faiza Sami, Muhammad Moeen Butt, Muhammad Amin
Summary: The two-parameter estimator (TPE) is proposed for the Poisson regression model, but it has the limitation of a single parameter. The count data models often suffer from the problems of dispersion and multicollinearity. The Conway-Maxwell-Poisson regression model (COMPRM) is suitable for handling both issues simultaneously. To estimate the COMPRM coefficients, the iterative reweighted least square (IRLS) method is used. Through a Monte Carlo simulation study, the efficiency of the estimator is evaluated based on the mean square error (MSE). In the presence of multicollinearity, the Asar and Genc's two-parameter estimator (AGTPE) shows better efficiency for COMPRM compared to other estimators like maximum likelihood (MLE), Ridge estimator, Liu estimator, and the TPE by Huang and Yang (HYTPE). The proposed estimator is also being studied for real-life applications.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2023)
Article
Engineering, Multidisciplinary
Asia Batool, Muhammad Amin, Ahmed Elhassanein
Summary: The Poisson Inverse Gaussian Regression model (PIGRM) is proposed to model count datasets and address over-dispersion. The maximum likelihood estimator (MLE) is commonly used for PIGRM, but it is not efficient when the explanatory variables are correlated. To overcome this issue, a ridge estimator is proposed and its properties are compared with MLE. Different ridge parameter estimators are also proposed. Simulation and real-life application results demonstrate the superiority of the proposed estimators over MLE using MSE as the performance evaluation criterion.
ALEXANDRIA ENGINEERING JOURNAL
(2023)
Article
Mathematics
Maryam Cheema, Muhammad Amin, Tahir Mahmood, Muhammad Faisal, Kamel Brahim, Ahmed Elhassanein
Summary: In statistical process control, control charts are used to monitor processes. This study proposes GLM-based control charts using different link functions and a binary response variable. Simulation study and real application on COVID-19 death monitoring demonstrate that the model-based control charts with the c-log-log link function perform better than other link functions.
Article
Mathematics, Applied
Muhammad Amin, Hajra Ashraf, Hassan S. Bakouch, Najla Qarmalah
Summary: This study proposes the James Stein Estimator for the beta regression model to address the issue of inaccurate estimation with correlated explanatory variables. Through simulation experiments and real-life applications, it is found that the proposed estimator outperforms other competitive estimators in estimating the parameters of the beta regression model.
Article
Mathematics, Applied
Muhammad Amin, Muhammad Nauman Akram, B. M. Golam Kibria, Huda M. M. Alshanbari, Nahid Fatima, Ahmed Elhassanein
Summary: This study proposed a new adjusted likelihood estimator for addressing the issue of multicollinearity in binary logistic regression models. Through Monte Carlo simulation and real application research, it was found that our proposed estimator performs better when there exists high but imperfect multicollinearity.
Article
Computer Science, Interdisciplinary Applications
Muhammad Amin, Azka Fatima, Muhammad Nauman Akram, Mustafa Kamal
Summary: This study compares the performance of different link functions in diagnosing influential observations in the logistic regression model. The results show that the CVR method with the logit link function is good for small explanatory variables. For large explanatory variables and small sample sizes, the cook's distance and DIFFITS with probit and logit link functions perform better than the CVR method. Similarly, for large explanatory variables and sample sizes, the cook's distance (with probit and logit link functions) and CVR with cauchit link function have the same performance and are better than the DFFITS method.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2023)
Article
Multidisciplinary Sciences
Faiza Sami, Muhammad Amin, Muhammad Moeen Butt, Seyab Yasin
Summary: The Poisson model, commonly used for counts type response variables, has a limitation of equal mean and variance compared to the Conway-Maxwell-Poisson-regression model (COMPRM). This study proposes an almost unbiased ridge estimator for the estimation of COMPRM coefficients and derives its theoretical properties. The results show that the proposed estimator outperforms the classical MLE and ridge estimators in terms of the minimum MSE and bias.
IRANIAN JOURNAL OF SCIENCE
(2023)
Article
Statistics & Probability
Muhammad Nauman Akram, Muhammad Amin, Nimra Afzal, B. M. Golam Kibria
Summary: The article discusses the parameter estimation for the zero-inflated negative binomial model in the presence of multicollinearity. A new estimator, called Kibria-Lukman estimator, is proposed along with some biasing parameters. The performance of the proposed estimator is compared with traditional biased estimators and found to be superior.
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
(2023)
Article
Statistics & Probability
Siddhartha Chakraborty, Biswabrata Pradhan
Summary: This paper studies some properties of weighted survival extropy measure, introduces weighted extended survival extropy and its dynamic version, and investigates various properties of the proposed measures. Generalized inequalities related to weighted extended survival extropy are also studied. Non-parametric estimators of these measures are proposed and their asymptotic properties are investigated. The performance of the estimators is assessed through a simulation study. Finally, it is shown that the weighted survival extropy measure can be used in model discrimination and as an alternative risk measure.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Xiaoliang Ling, Jiaojiao Zhang, Yinzhao Wei
Summary: This paper considers a reliability model for systems that fail due to degradation failure or catastrophic failure. The degradation process is modeled using a linear degradation path, and the catastrophic failure rate is dependent on the degradation rate. Firstly, stochastic comparisons are conducted between two systems with different stochastic orders for their respective degradation rate variables. Secondly, stochastic comparisons are conducted between two systems with different conditional catastrophic failure rates. Finally, examples are given using well-known models to demonstrate the validity of the results.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Jie Zhang, Yang Li, Ni Zhao, Zemin Zheng
Summary: This article discusses the widespread issue of corrupted data in many contemporary applications. It proposes a sparse modeling method based on L-0 regularization and efficiently solves the regularization problem using projection techniques. It proves the statistical properties of the proposed method under certain conditions and demonstrates its effectiveness through simulation studies.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Lin Zhu, Feifei Yan
Summary: An estimated quadratic inference function method is developed for the analysis of multivariate failure time data, where the primary covariates are incomplete but the auxiliary covariates for them are available for the whole cohort subjects. This method improves the estimation efficiency under the marginal hazard model with common baseline hazard function by incorporating both the auxiliary information and the intra-cluster correlation between the failure times. Simulation studies demonstrate that the proposed method gains noticeable efficiency compared to other existing methods when the intra-cluster correlation is strong or moderate.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Longxiang Fang, Shuai Zhang, Jinling Lu
Summary: This paper investigates the k-out-of-n system with n independent parallel subsystems comprising dependent components drawn from a heterogeneous population consisting of m different subpopulations. The components within each subpopulation are assumed to be dependent, while the subpopulations are independent of each other. The study discusses the reliability of the k-out-of-n systems by comparing two selection probabilities or two allocation policies in the sense of majorization order. Numerical examples are provided to illustrate the obtained results. Finally, some concluding remarks are made.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Amel Saidi, Abdelghani Hamaz, Ouerdia Arezki
Summary: This article presents new theoretical results on estimation in nonlinear random field models. The authors focus on a two dimensionally indexed random coefficients autoregressive model and develop a maximum likelihood estimation procedure for estimating the unknown parameters. The authors also prove the strong consistency of the estimates. The results are applied to construct efficient estimates in a specific model.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Qiang Zhang, Lijun Wu
Summary: In this article, a robust optimal proportional reinsurance and investment problem is considered in a model with delay and dependent risks. The insurer aims to develop the optimal strategy by maximizing the expected exponential utility of the combination of historical performance and terminal wealth. The expressions of the optimal strategy and value function are explicitly obtained using stochastic dynamic programming technique.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Jean-Claude Malela-Majika, Marien A. Graham
Summary: This study explores the idea of designing nonparametric charting schemes by using lower percentile points of the unconditional run-length distribution to restrict the probability of unconditional early false alarms. The research indicates that bias in the Phase-I sample may lead to remarkably high rates of early false alarms.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Letter
Statistics & Probability
Reza Farhadian
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)
Article
Statistics & Probability
Qiang Wu, Paul Vos
Summary: The article introduces the application of permutation inference distribution in multiple regression, which can be used for hypothesis testing and confidence interval construction with relatively low computational burden. The results show that PID confidence regions are asymptotically ellipsoidal and exact.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2024)