amir hossein amiri

Monitoring Binomial regression profiles in Phase II is examined in this study for multistage manufacturing processes where the quality characteristic is binary. In these kinds of processes, the quality of the final product depends on the quality characteristic of the previous stages, which is referred to as the cascade property. The U statistic was used to diminish the effect of this property. Then, four approaches, such as T2 and MEWMA control chart, LRT, and LRT/EWMA method, have been used, and the performance of these methods have been evaluated using simulation and a numerical example by means of ARL. An actual case study was also used to investigate the effectiveness of monitoring methods in further depth. Studies reveal that the proposed schemes perform well.

Many methods are applied to network surveillance for anomaly detection. Some quality control methods have been developed to monitor several quality characteristics simultaneously in different networks. In our study, we use three multivariate process monitoring techniques such as Hotelling’s T2, MEWMA, and MCUSUM to compare to the prior univariate control charts in the Degree-Corrected Stochastic Block Model (DCSBM), a random network model supporting the degree of each node based on Poisson distribution. By estimating parameters in Phase I from many charts, we apply ARL and SDRL metrics for the performance evaluation of multivariate control charts. The advantage of our method is detecting signals faster than previews ones by simulation and this is useful for defining the suitable method in different types of change. Furthermore, the quality of performance in different multivariate methods is displayed in detecting the shifts in the DCSBM. Finally, MCUSUM shows better performance for monitoring local and global changes than other methods.

In this paper, zero-inflated Poisson (ZIP) regression was assumed as an underlying model to generate network data. This model can be an appropriate model if the network data is sparse and produced with two processes, one generates only zeros and the other generates count data that follow the Poisson model, the two parameters of the model are functions of variables here referred to as similarity variables. The performance of the Likelihood Ratio Test (LRT), a Combined Residual-Square Residual (R-SR), and Hotelling's T^2 control charts was investigated in networks based on the ZIP regression model in Phase I. Traditionally, in Phase I the parameters of the model are unknown and need to be estimated. One needs to be sure the process is stable and the changes are detected and removed. The performance of our proposed methods is compared using simulation when parameters slope and intercept are under step changes. Signal probability was recorded as a comparison measure. The simulation results show that the LRT outperforms two other methods significantly in terms of signal probability. The efficiency of methods was also examined in real Enron data set

The quality of a process can be described using a regression profile relationship between a response variable and some independent variables. Much research has been done on the response variables with continuous and normal distribution. While, in real situations, when a product is conforming or nonconforming on the product line, the assumption of normality is violated and a logistic regression model is used to characterize binary response variables. Also, in many cases the parameters used to design control charts for monitoring profiles are unknown and estimated by using IC reference data, which adversely influences the efficiency of control charts. In recent years, a few authors have been done on the effect of parameter estimation in monitoring profiles, especially profiles whose response variables do not follow the normal distribution. In this paper, Hotelling’s T2 chart and a multivariate exponentially weighted moving average (MEWMA) chart are used to monitor the logistic regression profile in Phase II with estimated parameters. In addition, two criteria including average of average run length (AARL) and standard deviation of average run length (SDARL) are utilized to appraise the effect of parameters estimation in Phase I on the Phase II performance of designed control charts through simulation runs. The results illustrate that the performance of these charts is significantly affected by the estimated parameters in both IC and OC conditions. Also, two methods are utilized to decrease the effect of parameters estimation which include increasing the number of reference profiles in Phase I and modifying the control limits.

In some statistical processes monitoring (SPM) applications, relationship between two or more ordinal factors is shown by an ordinal contingency table (OCT) and it is described by the ordinal Log-linear model (OLLM). Newton-Raphson algorithm methods have also been used to estimate the parameters of the log-linear model. In this paper, the OLLM based processes is monitored using MR and likelihood ratio test (LRT) approaches in Phase I. Some simulation studies are applied to performance evaluation of the proposed approaches in terms of probability of signal under step shifts, drifts and the presence of outliers. Results show that, by imposing the small and moderate shifts in the ordinal log-linear model parameters, the MR statistic has better performance than LRT. In addition, a real case study in dissolution testing in pharmaceutical industry is employed to show the application of the proposed control charts in Phase I.

Control chart is one of the useful tools of statistical process control, which monitor the processes over time. In most of the designed control charts, it is assumed that the measurement errors do not exist in the measurement system, while this assumption is usually violated in practice as well. The presence of measurement errors leads to poor performance of the control charts. In this paper, a multivariate exponentially weighted moving average control chart is designed by considering measurement errors in Phase II. To decline the effect of measurement errors on the performance of the proposed control chart, multiple measurements method is applied. Also, sensitivity analysis about the effect of the number of measurements on the ARL performance of the proposed control chart is conducted. Note that different scenarios for the variance-covariance matrix are considered in simulation studies, including Case 1. Uncorrelated case with equal variances. Case 2. Negatively correlated case with equal variances. Case 3. Uncorrelated case with unequal variances. Case 4. Positively correlated case with unequal variances. Moreover, the performance of the proposed control chart is compared with the performance of Hotelling's T2 control chart. Results show the admissible performance of the proposed method in decreasing the effect of measurement errors.

In recent years, medical images have played an essential role in diagnosis, treatment, and training areas. Thus, any advancement in this field can help doctors in diagnosing. On the other hand, statistical process control (SPC) is now widely used in monitoring healthcare processes. In this research, using the image processing techniques and feature extraction methods (two-dimensional discrete wavelet), we propose some multivariate control charts to diagnose the type of bone marrow of the patients suspected of bone marrow metastasis in the pelvic region with early breast tumors. For this, 76 features (energy and histogram of oriented gradient) are extracted from the image. Next, using the GA, six features are selected and constitute a feature vector. Based on the feature vector, Hotelling’s T2 multivariate control charts are developed. Moreover, considering the high sensitivity of the classic estimators to outliers and contaminated data, we provide a robust Hotelling’s T2 control chart. Finally, we compare the ARL performance of the robust and the classic Hotelling’s T2 control charts in Phase II in the presence of local outliers in the Phase I data. The results confirmed the superiority of the robust version.

Statistical variables are divided into two categories: nominal and ordinal, both of which have many uses. In some statistical process monitoring applications, quality of a process or product is described by more than one ordinal quality characteristics called ordinal multivariate process. To show the relationship between these variables, an ordinal contingency table is used and modeled with ordinal log-linear model. In our manuscript, two new statistics including simple ordinal categorical and Generalized-p are developed for Phase II monitoring the ordinal log-linear model based processes. Performance of the proposed statistics are evaluated by using some simulation studies and a real numerical example. Results show the superiority of simple ordinal categorical based control chart. In addition, performance of these statistics is accessed through a sensitivity analysis on the size of the rows and columns of the contingency table. Meanwhile, a sensitivity analysis with three and four categorical factors is performed and similar results are obtained.

If the quality of a process is described using a linear functional relationship between the response variable and independent variables, such a relationship is called the profile. Today, with the development of manufacturing technologies, multistage processes have found a special position in manufacturing companies and industries. In this paper, we consider a multistage process with AR(1) auto-correlated simple linear profile in each stage and address the effect of both auto-correlation and cascade property on the performance of common monitoring procedures. To eliminate the effect of auto-correlation, we used a transformation method as a remedial measure at first. Then, an approach based on the U statistic is applied to remove the cascade property. Next, a modified T2 control chart is proposed to monitor the process in the second stage. The performance of the proposed control chart is evaluated in terms of average run length criterion. The simulation studies show that the proposed control chart perform satisfactorily.

Independency of observations is one of the fundamental assumptions in control charts. However, in some processes this assumption is violated and data are auto-correlated. Also, it is assumed that the measurement errors are absent in measurement system while, this assumption is usually violated. The existence of the auto-correlation and measurement errors causes the poor performance of the control charts. In other words, the average run length in the case of out-of-control(OC) situations increases in the presence of auto-correlation and measurement errors. In this paper, the effect of auto-correlation and measurement errors on the performance of Hotelling’s T2 control charts in Phase II in multivariate normal processes is investigated in terms of average run length(ARL) criterion. The first order auto-regressive model as auto-correlation structure between observations within each sample is discussed in this paper. To decrease the effect of auto-correlation and measurement errors on the performance of the Hotelling’s T2 control chart, jump strategy and multiple measurements methods are applied, respectively. The effect of auto-correlation and measurement errors, individually and simultaneously, as well as the performance of the suggested methods to address these effects is appraised through simulation studies and a numerical example. The effect of number of measurements and jumps on the ARL values of the proposed control chart is also evaluated. Results show the acceptable performance of the multiple measurements and jumps methods in diminishing the effect of measurement errors and auto-correlation, respectively. At last, a real case is presented to show the application of the proposed scheme.

In this paper, the economic-statistical design of the Max EWMAMS control chart under measurement errors and multiple measurements for joint monitoring of mean and variability of the process is investigated. The traditional approach for monitoring mean and variance of the quality characteristic is using two separate control charts. This approach leads to an increase in the probability of Type I error. To overcome this problem, researchers have proposed control charts for joint monitoring of mean and variability of the process. Also in practice, the measurement errors exist in the sampling process. The sampling with multiple measurements is a way to reduce the deficiency of the measurement error and increasing power of the control chart. However, the multiple measurements cause to increase the sampling costs. Hence, this factor should be considered in the economic-statistical design of control charts as well. In the proposed cost model, the Lorenzen-Vance cost function is developed and a genetic algorithm is applied to obtain model parameters that minimize cost function. Finally, the performance of the proposed model is evaluated by a numerical example.

Nowadays, the optimal use of energy, speed increase, fuel consumption reduction and environmental pollution prevention are essential in the ship design and manufacturing discussion. So it is critical to be able to estimate ship's response to waves, since the resulting added resistance and increasing fuel consumption also increasing environmental pollution. In this study, the numerical solution of the Iran-Bushehr container ship model was determined to simulate the free surface flow around the hull and estimate the total resistance and investigate ship motions in 2 degree-of-freedom (heave and pitch) in regular head waves. Finally, the results of simulation in different conditions are presented and compared. KCS container ship used for validation. CFD codes in Star-CCM+ used to calculate three dimensional, incompressible, unsteady RANS equations.

One of the currently important and widely used research subjects in the healthcare area of cancer patients is the diagnosis procedure of cancer tumors and metastases in magnetic resonance imaging such that it has a high level of accuracy and also be a support for doctors in interpreting and diagnosing medical data. To this aim, a multivariate Hotelling’s T2 control chart is used.

Using a two-dimensional discrete wavelet transform, some features of the image texture are extracted by using statistical and transform methods. Then, to reduce the data dimensions and feature selection, a genetic algorithm is used. Afterward, two methods including fuzzy c-Means clustering algorithm and a multivariate Hotelling’s T2 control chart are used to diagnose bone marrow metastasis patients.

From 204 bone marrow samples, 76 features are extracted from which six ones are selected and a 204×6 feature vector matrix is generated. Finally, the performance of the proposed two methods is compared. The results show that the diagnosis and accuracy measures of multivariate Hotelling’s T2 control chart are better than the other method.

In the context of cancer, one of the current concerns for healthcare providers is to use non-invasive, short response time, and highly accurate methods in diagnosing tumors and metastases. The proposed method appropriately addresses these requirements.

In some applications of statistical process monitoring, a quality characteristic can be characterized by linear regressionrelationships between several response variables and one explanatory variable, which is referred to as a “multivariate simplelinear profile.” It is usually assumed that the process parameters are known in Phase II. However, in most applications,this assumption is violated; the parameters are unknown and should be estimated based on historical data sets in Phase I.This study aims to compare the effect of parameter estimation on the performance of three Phase II approaches for monitoringmultivariate simple linear profiles, designated as MEWMA, MEWMA_3 and MEWMA∕2 . Three metrics are usedto accomplish this objective AARL, SDARL and CVARL. The superior method may be different in terms of the AARL andSDARL metrics. Using the CVARL metric helps practitioners make reliable decisions. The comparisons are carried outunder both in-control and out-of-control conditions for all competing approaches. The corrected limits are also obtained bya Monte Carlo simulation in order to decrease the required number of Phase I samples for parameter estimation. The resultsreveal that parameter estimation strongly affects the in-control and out-of-control performance of monitoring approaches,and a large number of Phase I samples are needed to achieve a parameter estimation that is close to the known parameters.The simulation results show that the MEWMA and MEWMA∕2 methods perform better than the MEWMA_3 method interms of the CVARL metric. However, the superior approach is different in terms of AARL and SDARL.

An essential tool for examining the quality of manufactured products is acceptance sampling. This research applies the concept of minimum angle method to extend two variables sampling plans including the variables multiple dependent state (VMDS) sampling plan and the variables repetitive group sampling (VRGS) plan on the basis of the process yield index Spk. Optimal parameters of acceptance sampling plans can be determined by solving a non-linear optimization model with the following conditions: 1) The objective function of the plan is to minimize the average sample number. 2) Constraints are set in a way that the compliance rate will be satisfied with the ideal operating characteristic (OC) curve as well as the producer’s and costumer’s risks. The assessment of the proposed plans reveals that by increasing the rate of convergence to the ideal OC curve, the proposed VRGS plan performs better than the proposed VMDS plan in terms of the average sample number. A numerical example is considered to reveal the applicability of the proposed acceptance sampling plans.

In some statistical process monitoring applications, quality of a process or product is described by more than one ordinal factors called ordinal multivariate process. To show the relationship between these factors, an ordinal contingency table is used and modeled with ordinal log-linear model. In this paper, a new control charts based on ordinal-normal statistic is developed to monitor the ordinal log-linear model based processes in Phase II. Performance of the proposed control chart is evaluated through simulation studies and a real numerical example. In addition, to show the efficiency of ordinal-normal control chart, performance of the proposed control chart is compared with an existing Generalized-p </em>chart. Results show the better performance of the proposed control chart in detecting the out-of-control condition.

In some profile monitoring applications, the independency assumption of consecutive binary response values within each profile is violated. To the best of our knowledge, estimating the time of a change in the parameters of an autocorrelated binary profile is neglected in the literature. In this paper, two maximum likelihood estimators are proposed to estimate the real time of step changes and drift in Phase II monitoring of binary profiles in the case of within-profile autocorrelation, respectively. Our proposed estimators, not only identify the change point in the autocorrelated logistic regression parameters, but also in autocorrelation coefficient. The performance of the proposed estimators to identify the time of change points either in regression parameters or autocorrelation coefficient is evaluated through simulation studies. The results in terms of the accuracy and precision criteria show the satisfactory performance of the proposed estimators under both step changes and drift. Moreover, a numerical example is given to illustrate the application of the proposed estimators.

In this paper, a new robust method based on weighted maximum likelihood estimation (WMLE) is proposed to estimate the regression parameters in logistic profiles in Phase I. This approach reduces the outlier’s effects on the statistical performance of T2 control chart in terms of probability of Type I error. A numerical example is used to evaluate the performance of the proposed method. The results show the better performance of the proposed estimator compared to the maximum likelihood estimation method in terms of power in T2 control chart.

Missing observations occur in experimental designs as a result of insufficient sampling, machine breakdown, high cost, and errors in the measurements. In nanomanufacturing, missing observations often appear in designs because the combination of factors or molecular structures selected by a designer cannot be experimented successfully. In the current paper, Box-Behnken and face-centered composite designs were studied and eight robustness criteria including D-efficiency, tmax, tmaxþ(), and their related sub-criteria were considered to evaluate the robustness of the aforementioned designs. Finally, the integrated TOPSIS-AHP methodology was employed to select the most suitable robust design, and a numerical example was also presented to assess the applicability of the proposed approach

Planning of useful and sustainable strategies is one of the most important goals of organizations to defend critical
systems. In this research, a modeling is considered for investment optimization of defense and attack in
complex with interdependent subsystems, in which failure of a subsystem will possibly affect the optimal performance
of other subsystems. In this study, a static model is proposed that according to the probabilities of a
successful attack, subsystems dependency ratio, different modes of operation of the system, reliability structure
and game theory approach in determining balancing point, presents a nonlinear planning model to determine the
amount of investment in defending and attacking of all subsystems. Then, according to the results obtained from
the proposed static model, the dynamics of the system and the concepts of evolutionary game theory, a new and
dynamic method is introduced to determine the stable strategies for defense and attack. According to the proposed
model, the evolutionarily stable strategy will be examined over time, from the perspective of a defender, attacker,
and the whole system. Finally, the proposed model is applied to a numerical example and its results are analyzed

In some processes, quality of a product should be characterized by functional relationships between response variables and a signal factor. Hence the traditional methods cannot be used to find the optimum solution. In this paper, we propose a method which considers two different dispersion effects, i.e. in domain and between replicates variations in the functional responses. Besides, we propose an integral based measure to find the deviation from target function. A probabilistic method is applied to consider the correlation structure of functional responses. Three numerical examples and a real case from literature are studied to show the efficiency of the proposed method

There is considerable interest in the use of Statistical Process Control (SPC) in healthcare in the recent years because SPC tools lead to continual improvement in healthcare process. Control chart is one of the main tools of the SPC, however, they usually signal the out-of-control status with delay respect to real time of the change known as change point. Hence, change point estimation is important in healthcare with considering relationship between quality engineering and hospital epidemiology. There are varieties of quality characteristics in healthcare which are should be monitored over time. Therefore, in this paper, first, g and h control charts are described since these control charts are famous tools for monitoring events in healthcare. Then, corresponding step change point estimators using Maximum Likelihood Estimation (MLE) are proposed. In this regard, Mont Carlo simulation is used to evaluate performances of the proposed estimators based on accuracy and precision measures under all kinds of shifts. In addition, cardinality and coverage probability of confidence set are presented for the proposed estimators based on the logarithm of the likelihood function.
The simulation studies are conducted under different magnitudes of the step shifts to evaluate performances of the proposed change point estimators in both control charts. The results show that as the magnitude of the step change in the parameter of the distributions increase, the performance of the proposed change point estimators improve significantly in terms of precision and accuracy measures. Also, cardinality and coverage percentage of confidence set estimators are calculated and plotted to show the relationship among these measures and increasing and decreasing shifts under the given reference values. In addition, the simulation studies demonstrate that as the magnitude of the step change in the parameters of the distributions enlarge, cardinality of set estimators reduces and coverage percentage of confidence set estimators increases under the given references values. In general, results show that the proposed change point estimators perform satisfactory under all types of shifts. Finally, the performance of one of the proposed change point estimators is illustrated through an applied example in healthcare.

Earned value management (EVM) is a well-known approach in a project control system which uses some indices to track schedule and cost performance of a project. In this paper, a new statistical framework based on self-starting monitoring and change point estimation is proposed to monitor correlated EVM indices which are usually auto-correlated over time and non-normally distributed. Also, a new change point estimator is developed to find the real time of change in the indices mean. Furthermore, a new diagnosing method is presented to recognize the deviated mean index. The performance of the proposed methods is evaluated through simulation studies and an illustrative example.

In some application, quality of product or performance of a process described by some functional relationships between some variables known as multivariate linear profile in the literature. In this paper, we propose Max-MEWMA and Max-MCUSUM control charts for simultaneous monitoring of mean vector and covariance matrix in multivariate multiple linear regression profiles in Phase II. The proposed control charts also have ability to diagnose either the location or variation of the process is responsible for out-of-control signal. The performance of the proposed control charts is compared with existing method through Monte-Carlo simulations. Finally, the applicability of the proposed control charts is illustrated using a real case of calibration application in the automotive industry.

Today, the crisis and its consequences, is one of the main threatening factors in organizations, and to determine efficient and sustainable strategies, according to the terms of defense and attack, proper planning must be done. In this study, with the aim of improving reliability, first, modeling the optimal strategies to defend and attack in the stationary state is presented, provided that the attacker to deceive the defender will create a number of false attacks. In the static model, considering the probability of a successful attack, defender capability in identifying false attacks, reliability block diagram and game theory approach in finding the balance point, a nonlinear programming model is proposed to determine the amount of investment defend and attack. Then, according to the results of the static model, system dynamics and implications of evolutionary game theory, a new and dynamic approach to determine sustainability strategies of defense and attack is presented, that defender use two strategies based on actual or 90% of all adopted attacks and the attacker also use two strategies use or not to use false attacks. Finally, presented model is illustrated for an applied case and final findings are analyzed.