M.Phil / MS

Permanent URI for this collectionhttps://repository.cuilahore.edu.pk/handle/123456789/63

This collection archives the complete set of theses produced by students of the COMSATS University Islamabad, Lahore Campus.

Browse

Now showing 1 - 20 of 20

A Four-Parameter Generalized Burr Distribution and Its Applications
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Sadia Sarfraz Khan; SP17-RST-002; LHR TP 5325; Dr Muhammad Mohsin
We study a new distribution called “A Four Parameter Generalized Burr Distribution” and present a detailed account of some of its structural properties including the r-th moment, moments about mean, skewness, kurtosis, and order statistics. The parameters of the under study distribution are estimated by the maximum likelihood method. The simulation study is also carried out under for different sample sizes to test the effectiveness of this model. Finally, the performance of the model is checked by using the real data set and comparing it with other known distributions.
Bayesian Estimation of Inverse Rayleigh Distribution using Truncated Inverted Gamma as a Prior with different Loss Functions
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2019) Sehrish Liaquat; FA17-RST-003; LHR TP 5744; Dr. Muhammad Mohsin
We introduce the truncated prior in Bayesian analysis to refine the posterior distribution and get improved results. The parameter of Inverse Rayleigh (IR) distribution is estimated through Bayesian approach using various loss functions such as squared error loss function, entropy, general entropy, precautionary and squared log error loss function. Two types of Bayes estimators of the posterior expected risks for each loss function i.e. the truncated prior with the proper sampling distribution and the prior as a proper pdf with truncated sampling distribution are studied. In addition, the equal-tail credible intervals are constructed for all the cases. The simulation study is conducted to compare the different Bayes estimators along with the different procedures to seek the best estimators. A real life example is also presented to support the performance of the proposed technique.
Data-Driven Sample Size Adaptation in EWMA Control Chart for Improved Monitoring
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2025) Gulmakai; CIIT/FA23-RST-002/LHR; Dr. Muhammad Noor-ul-Amin; LHR TP 9873
Statistical Process Control (SPC) serves as a fundamental methodology for monitoring and maintaining stability in manufacturing processes. In this study, an Adaptive Sample Size-based Exponentially Weighted Moving Average (ASEWMA) control chart is proposed to enhance the detection of process shifts by dynamically adjusting the sample size in response to observed process behavior. The adaptability of the sample size enables the control chart to respond more efficiently to both small and moderate shifts while conserving computational and sampling resources when the process is stable. Further, the proposed control chart is evaluated in the presence of measurement error. The performance of the ASEWMA control chart is rigorously evaluated through extensive Monte Carlo simulations. Key performance indicators such as the Average Run Length (ARL) and Standard Deviation of Run Length (SDRL) are employed to assess its effectiveness in various shift scenarios. Comparative analyses demonstrate that the ASEWMA chart consistently outperforms traditional control charts, including the Fixed Sample Size EWMA (FEWMA) and Variable Sample Size EWMA (VEWMA), particularly in detecting small shifts in the process mean. The findings of this research highlight the ASEWMA chart’s potential to provide a practical and efficient tool for quality monitoring in industrial settings. By achieving a desirable balance between sensitivity to shifts and computational efficiency, the proposed method facilitates the timely detection of process changes while maintaining robustness under in-control conditions
Development of Estimators for Population Mean of Sensitive Variable in Survey Sampling
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2017) Muhammad Kamran ASLAM; SP16-RST-005; LHR TP 4833; Dr. Riffat Jabeen
In Chapter 1, introduction about Survey sampling, use of auxiliary in formation in context stratified random sampling has been deliberated. Further in Chapter 1, Different methods dealing randomized response have been deliberated by different authors but Hansen-Hurwitz sub sampling has been consider in more detail. In Chapter 2, the literature regarding the use of classical sampling de sign has been deliberated. The literature about some existing estimators for calibration estimator under stratified random sampling has been ex press. Various work of randomized response has also been specified in Chapter 2. Some well-known existing randomized response model, scrambling vari able under auxiliary and study variable, with randomized response or complete response have been reproduced in Chapter3. The major con tribution of this study starts from next Chapter. In Chapter 4, new estimators using auxiliary information have been proposed for calibration estimator in randomized response under strat ified random sampling scheme. The mean square error each estimator has been derived along. In Chapter 5 conclusion has been drawn about the generalized class of estimators proposed in this study. Finally it is shown that the propose estimator II is more efficient than proposed estimator I.
Development of Fuzzy X ̅-S Control Charts through incorporating Fuzziness in Conventional X ̅-S Control Charts
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Nabila Ali; SP17-RST-006; LHR TP 5324; Dr. Tajammal Hussain
The mission to progress the quality and productivity of products and services is spreading across the industry and services sectors. Statistical process control (SPC) is a methodology to monitor an ongoing process and to identify unusual patterns and to indicate the need for correction. The most common SPC tools to be used are control charts. Control charts indicate whether or not the process is stable and resulting quality improvement. Control charts help to distinguish between special and common causes of variation which may occur during manufacturing process and influence the quality of a product.
Development of Modified MaxAEWMA Control Chart based on Auxiliary Information
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2021) Zainab Manzoor; SP19-RST-004; LHR TP 6536; Dr. Tajammal Hussain
Globalizations and streamlined commerce have opened the entryway of worldwide open market rivalry that changed the organization of working together. Open rivalry has an extreme impact on neighborhood market makers and worldwide firms. From the earliest starting point, quality is considered as a pre-essential definitive component to choose items or administrations. Presently days, the associations put more into observing and improving the nature of the items to draw in more clients. Besides, the basic considering purchasers has changed now they request a higher caliber of items and administrations. Walter A. Shewhart (1920) presented the idea of Statistical Quality Control (SQC), which is a coordinator of the quality control field. The crucial target of SQC is to zero in on the nature of the assembling mechanism to accomplish the most noteworthy creation quality. As per Montgomery (2009) in statistical mechanism control, each made item is not the same as one and another due to normal inborn variety, however with unnoticeable variety. The vast majority of organizations are putting resources into the quality enhancements division to decrease squandered items. The decrease in changeability builds the benefit of association and fulfillment of clients by satisfying their needs. As observed from the shopper's conduct a high variety of items is unsuitable for them, consequently, the objective of associations is to create items by stable mechanism.
EBIZ Lounge Private Limited
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2025) Sehar Shabbir; CIIT/FA21-BST-011/LHR; Dr. Tajammul Hussain; LHR TP 9931
EBIZ LOUNGE (PRIVATE) LIMITED, a well-established service company with over two decades of experience, has built a strong reputation in the industry in Pakistan by consistently delivering high-quality services to its clients. The report includes an introduction to the company, along with an overview of its organizational structure, competitive strategy, and core business operations. A brief SWOT analysis is also provided, highlighting the strengths, weaknesses, opportunities, and threats facing the organization. The final section covers the responsibilities that were assigned, the knowledge gained from the experience, and some of the challenges encountered during the internship. Through the internship at EBIZ LOUNGE (PRIVATE) LIMITED, valuable exposure to the exports industry was gained. The insights and expertise acquired throughout the experience added to the professional growth, providing the opportunity to apply the knowledge learned through education.
Estimation of Mean for a Finite Population Using Sub Sampling of Non-Responden
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Iqra Ehsan; SP16-RST-004; LHR TP 4834; Dr. Aamir Sanaullah
In Chapter 1, introduction about Survey sampling, use of auxiliary information in context of single and two-phase sampling has been deliberated. Further in Chapter 1, the problem of non-response in two has been deliberated. Different methods dealing non-response have been deliberated by different authors but Hansen-Hurwitz sub-sampling has been deliberated in more detail. In Chapter 2, the literature regarding the use of classical sampling design has been deliberated. The literature about some existing estimators for single and two-phase sampling has been deliberated. Various work of non-response has also been specified in Chapter 2. Research methodology of proposed work has been given in chapter # 3. The major contribution of this study starts from Chapter 4 & 5. In chapter 4, we proposed some estimators using single auxiliary variables when population mean of auxiliary variable is available under simple random sampling in the occurrence of non- response. Further the properties of the proposed class of estimators have been deliberated in chapter 4. Also optimum conditions for which proposed class of estimators has minimum MSE’s has been deliberated with their empirical studies. In chapter 5, we proposed some estimators using single auxiliary variable when population mean of auxiliary variable is not available in prior under simple random sampling in the occurrence of non- response. Further the properties of the proposed class of estimators have been deliberated in chapter 5. Also optimal conditions for which proposed class of estimators attain minimum MSE’s has been deliberated with their empirical studies. In Chapter 6 conclusion has been drawn about the generalized class of estimators proposed in this study. Finally it is shown that the proposed estimators are more efficient as compare to the existing estimators of literature
Estimation of Population Mean by Using Robust Regression
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Salah ud din Asghar; SP16-RST-003; LHR TP 5250; Dr. Muhammad Noor-ul-amin
In the linear regression problems, ordinary least square (OLS) estimates violate in the presence of outliers as according to the assumptions of the OLS model that there should not be outliers in the data. More specifically, since the classical procedures are not applicable to the case of outliers presence. So, for the purpose of estimation there is another approach which handles this situation of outliers that is robust regression. Robust regression is an important tool for investigating data that are tainted with outliers and provide efficient results. In robust regression, estimates completely reject the effect of outliers and provide good estimates of the population parameter. In this work we focus on the problem of outliers occur in the data which provides a bad estimate of the population mean. In this dissertation, redescending M-estimator and ratio estimator have been proposed to estimate finite population mean using the information from auxiliary variable. Discussions have been made about the robust estimation method based on M-estimation, outliers, ratio estimators and auxiliary variable in Chapter 1. The literature about the robust regression and ratio estimator is discussed in Chapter 2. This research work is based on the two main chapters. The 3rd and 4th Chapters are the most important chapters of the thesis. For the first strategy, in the 3rd Chapter a redescending M-estimator is proposed by reworking on the redescending M-estimator proposed by Insha-ullah et al. (2006) which assign a weight closer to zero to the extreme values and closer to one to the good observations. Most of the redescending M-estimators are solved by iteratively reweighted least square method. For the second strategy we adopt the ratio estimators proposed by Kadilar and Cingi (2004) in the 4th Chapter and replace OLS coefficient with our redescending M-estimator proposed in 3rd Chapter in these ratio estimators. In both parts numerical illustrations and simulation studies were used to analyze the results of the proposed estimators with the comparative estimators to achieve the main results of the research. The intention of this research is to analyze, introduce and improve robust estimation methodologies to solve real life problems such as outliers in data. And finally in 5th chapter conclusion is written based upon whole the results of research.
Estimation of Population Mean of a Sensitive Variable under Systematic Sampling
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Muhammad Kamran Khan; FA16-RST-002; LHR TP 5310
In this research study, we consider the problem of mean estimation of a sensitive variable under systematic sampling. The estimator based on one auxiliary variable is proposed for the sensitive variable under systematic sampling. Further, estimators are studied in the presence of non-response. The mean square error (MSE) expressions are derived and efficiency of proposed estimator is proved by conducting simulations.
Generalized Inverted Exponential Distribution: Theory and Applications
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2025) Muhammad Umar Amin; CIIT/FA22-RST-005/LHR; Dr. Mian Muhammad Farooq; LHR TP 9871
In this research work we introduced generalization of the Inverted Exponential Distribution based on survival function. Our aim is to provide an extensive treatment of the mathematical properties of new model. The possible extension of corresponding probability density and distribution function have been derived and present. Some distributional properties of new model are discussed and order statistics have been derived. Maximum likelihood estimation (MLE) is sued to estimate the model parameters and real-life data application have also used to study the proposed model. Additionally, the study uses Bayesian estimation techniques, obtaining posterior estimates of the model parameters through the use of the Markov Chain Monte Carlo (MCMC) procedure. This method offers more accurate inference than the standard methods. Furthermore, Type-II censoring is used to evaluate the model performance, indicating its flexibility and reliability.
Generalized Ratio Type Exponential Estimators for the Population Mean Using Ranked Set Sampling
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Amjad Ali; SP16-RST-002; LHR TP 4837; Dr. Muhammad Ismail
The dominance of using ranked set sampling, for estimating the mean of a population, as compare to simple random sampling is well reputable. This technique is useful when visual ordering of a small set of size “m” can be made simply and precisely, since exact measurement of a value is most of the times very hard to obtain. In this dissertation, exponential ratio type estimators are proposed for evaluating the population mean by using ranked set sampling. We found through numerical results that the results of proposed ratio type estimators which is estimated under ranked set sampling is more reliable and competent than the simple random sampling (SRS) and some other already existed estimators under SRS are suggested by Cochran (1977), Parsad 1989, Bahl and Tuteja (1991). The comparison is also made with estimators under ranked set sampling by Samawi and Muttalak (1996), Rabail (2014) and Gajendra , Zeeshan and Carlos (2017).
Hybrid Exponentially Weighted Moving Average Control Chart Using Ranked Set Sampling
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Navara Shabbir; SP17-RST-001; LHR TP 5301; Dr. Muhammad Noor-ul-Amin
In manufacturing process, the products are subjected to variations which directly impact the quality of a product e.g. in the process of filling juice bottles, the amount of juice filled may not exactly the same. Quality control combined with statistical techniques is called statistical process control (SPC). To achieve the level of perfection in a production procedure SPC play an important role, it generally contains tools for monitoring the dissimilarities caused during a production process. SPC has a direct impact in increasing the quality of a product by reducing the amount of variation and enabling it to attain desired satisfaction level.
Induction of Measurement Error in Acceptance Sampling Plans
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2019) Sajid Ali Naqvi; SP17-RST-004; LHR TP 5741; Dr. Muhammad Noor-ul-Amin
Measurement error practically exists to a certain degree and part and parcel of measurement process for quantitative observations. Measurement error cause contamination and result in ineffectiveness of the scale sampling plan. In present study the effect of measurement error is investigated for EWMA based scale sampling plan using a linear covariate method when population standard deviation is known or unknown. The same effect is also examined in the presence of auxiliary information when measurement of quality characteristic is not feasible or accessible. Multiple measurements method is a remedy to minimize the effect of measurement error. The variance of measurement error component becomes zero for infinite number of multiple measurements. A reasonable and economical number of multiple measurements may restore the effectiveness and utilization of EWMA based scale sampling plans with or without auxiliary information when population standard deviation is known or unknown. Tables are provided for various values of model parameters for industrial use
On Extended Exponential Weighted Moving Average Control Charts for Monitoring Mean of a Process
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2020) Aqsa Hussain; SP19-RST-001; LHR TP 6535; Dr. Aamir Sanaullah
In chapter 1, the introduction about statistical process control, control charts and use of auxiliary information in control charts has been explained. Further in chapter 1, objectives of control charts also described. Objectives of the proposed study explained in chapter 1. In chapter 2, literature about some existing EWMA-type control charts has been provided. Various works of EWMA-type control charts using auxiliary information have also been specified in chapter 2. In chapter 3, an extended-EWMA control chart by Naveed et al. (2018) has been discussed. Further a EWMA control chart based on the auxiliary variable due to Abbas et al. (2014) for mean has been deliberated. In chapter 4, we are working on EEWMA control chart by using an auxiliary information for improving the efficiency of an extended-EWMA control chart due to Naveed et al. (2018), and then comparison of the proposed EEWMA with the existing EEWMA and EWMA control charts is presented. In chapter 5, we propose EEWMA sub-group mean control charts for mean with or without using auxiliary information to detect smaller shifts earlier. Comparison with the previous control charts is also done in this chapter. Finally it is shown that the proposed EEWMA control charts are more efficient as compare to the existing control charts of literature
Revisiting the Process Capability Indices (PCIs) for the Uni-variate Variable Control Charts by Incorporating the Concept of Fuzziness
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2019) Muhammad Usman Aslam; FA17-RST-001; LHR TP 5742; Dr. Tajammal Hussain
With the passage of time and global development in the industrial world, the competition for the quality improvement of the manufactured product has also increased. To achieve the quality perfection goal the best utilized technique is quality control. The technique of quality control is merged with the statistical methods to obtain a reliable source for updating and improving the product’s quality and services, it also helps to maintain the actual quality standard. This combination of quality control with the statistical methods is called Statistical Quality Control (SQC). The concept of Statistical Quality Control (SQC) was introduced in 1924 by Dr. Shewhart who is well known as the father of quality control analysis. The principle objective of SQC is to focus on the quality of any organization and to achieve the optimum quality of production as well as services by using the sufficient statistical techniques. According to (Montgomery, 2009) any products quality depends upon its fittness for use and it is judged by the appropriate quality characteristics.
Some Memory Type Estimators for Population Variance in Survey Sampling
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2025) Umme Habiba; CIIT/FA23/RST/004/LHR; Dr. Riffat Jabeen; LHR TP 9874
Reducing and estimating population variation is important in survey sampling. These variations can occur in any sampling design, including stratified random sampling. In stratified random sampling the difference in stratum weights increase the variation. This problem can be control with the use of calibration techniques, and auxiliary information. By using this technique, we can increase the accuracy and efficiency of estimator. Improving the efficiency of estimator for population variance estimation is the primary goal of this study. Both exponentially weighted moving average (EWMA) and extended exponentially weighted moving average (EEWMA) memory type statistics are used in this study to estimate population variance. EWMA use only current data whereas EEWMA use both current and past observations. The EWMA and EEWMA statistics are used to build the calibration variance and ratio estimators. The mean square errors are calculated, and the effectiveness of the estimators is illustrated through a discussion of mathematical comparisons. To assess the efficiency of the proposed memory type estimators, a simulation study is conducted and MSE are compare with existing estimators. The results show that proposed estimators are perform efficiently as compared to the existing estimators
Some New Insights of Exponential Weighted Moving Average Control Chart with Fuzzy Control Lenses
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2018) Muhammad Javed; SP16-RST-001; LHR TP 5249; Dr. Tajammal Hussain
Statistical process control (SPC) is an approaching path that analyzes processes despite the fact whether they are under statistical control or out of control. So, Control charts are commonly employed for this purpose. Possibly, sample data may possess some uncertainties occurring owing to systems of measurement and conditions of the surrounding environment, for this purpose fuzzy numbers or language variables are usable to seize such form of uncertainties. Thereby, the well-known control charts, exponentially weighted moving average control chart (EWMA) to univariate data are formed up under the environment that is fuzzy one. The fuzzy EWMA control charts (FEWMA) are also usable in order to detect minute changes in the data presented by fuzzy numbers. FEWMA lessens various mistaken and unreliable decisions by catering further flexibleness upon the control limits. As far as the production process is monitored and controlled by dint of FEWMA control chart.
Some Properties, Characterization and an Application of Mittag-Leffler type Poisson distribution
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2020) AYESHA ISRAR; SP19-RST-006; LHR TP 6537; Dr. Muhammad Mohsin
Probability distributions analyze the complex phenomena in a comprehensive way and provide the possible solutions of the understudy problem. The characteristics of a probability distribution describe its behavior and check its compatibility for modeling the data. The present research involves an extensive study of Mittag-Leffler type Poisson (MLTP) distribution developed by Porwal and Dixit (2017). They used Mittag-Leffler function, a mathematical function/series, to develop a new probability distribution. This function has been extensively used in the fields of mathematics, engineering, medical and social sciences. Some mathematical properties of the MLTP distribution are discussed. Graphical representation of MLTP distribution is provided for different values of the parameters. The moment generating function, characteristic function and order statistic are also studied. Characterizations for the sum of the random variables are presented. The distributions of product and ratio of two MLTP random variables are derived. In addition, the estimates of unknown parameters of MLTP distribution are obtained by using the method of maximum likelihood estimation. The MLTP distribution is applied to three real life data sets from the medical field. The log-likelihood values, AIC and BIC of the present model are compared with those of some other distributions extant in the literature.
Uncovering the Statistical Foundations of Bibliometric Techniques and Their Role in Exploring the Data Science Research Evolution
(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2025) Aleem Ahmad; CIIT/FA23-RST-001/LHR; Dr. Tajammal Hussain; LHR TP 9872
Sceinometric and bibliometric techniques are widely used in assessing the quality and quantity of research production in all aspects. In sceinometrics techniques, we focus on evaluating the scientific progress in all fields of science, while in bibliometrics, which is basically is a subpart of sceinometrics, we mainly focus on the quality and quantity of the research production. Both techniques are very important for finding the current research trends. These techniques are widely used by many research-funded institutions, decision makers. and the government. These help them with research funding allocations and identify the emerging research areas. These techniques are very useful in evaluating the journal impact factor, ranking universities, countries, and enhancing research efficiency. Because all research production activity creates data so these techniques are very closely related to data science. In Data Science, we collect, analyze, and process the largest sets of research data. We can find its application in every field, such as healthcare, education, and environmental studies, etc. In this study, we are focusing on the statistical foundation of bibliometric techniques and see that what is their role in exploring the data science research field. Many statistical models, such as discriminant analysis, cluster analysis, and VOSviewer software, are used for viewing the scientific landscape of this research. The present research gives us in-depth insight into the progression and evolution of data science as a prominent research field, after using advanced bibliometric techniques on the dataset of all 4424 journal articles till 31 December 2023. This study concludes that data science is evolving and has a potential research domain.