Simulated maximum likelihood estimation in joint models for multiple longitudinal markers and recurrent events of multiple types,in the presence of a terminal event

In medical studies we are often confronted with complex longitudinal data. During the follow-up period, which can be ended prematurely by a terminal event (e.g. death), a subject can experience recurrent events of multiple types. In addition, we collect repeated measurements from multiple markers. An adverse health status, represented by ‘bad’ marker values and an abnormal number of recurrent events, is often associated with the risk of experiencing the terminal event. In this situation, the missingness of the data is not at random and, to avoid bias, it is necessary to model all data simultaneously using a joint model. The correlations between the repeated observations of a marker or an event type within an individual are captured by normally distributed random effects. Because the joint likelihood contains an analytically intractable integral, Bayesian approaches or quadrature approximation techniques are necessary to evaluate the likelihood. However, when the number of recurrent event types and markers is large, the dimensionality of the integral is high and these methods are too computationally expensive. As an alternative, we propose a simulated maximum-likelihood approach based on quasi-Monte Carlo integration to evaluate the likelihood of joint models with multiple recurrent event types and markers.     

Estimation of parameters of Misclassified Size Biased Discrete Lindley Distribution

In this paper, a two-parameter discrete distribution named Misclassified Size Biased Discrete Lindley distribution is defined under the situation of misclassification where some of the observations corresponding to x = c + 1 are reported as x = c with misclassification errorα. Different estimation methods like maximum likelihood estimation, moment estimation, and Bayes Estimation are considered to estimate the parameters of Misclassified Size Biased Discrete Lindley distribution. These methods are compared by using mean square error through simulation study with varying sample sizes. Further general form of factorial moment is also obtained for Misclassified Size Biased Discrete Lindley distribution. Real life data set is used to fit Misclassified Size Biased Discrete Lindley distribution.     

The effect of the mechanism and amount of missingness on subset correspondence analysis

The application of subset correspondence analysis is a relatively new technique to deal with the analysis of categorical data with missingness. A simulation study is used to test the effects of Little and Rubin's missingness mechanisms, as well as missingness up to 50% on subset correspondence analysis. Missingness was simulated across 18 different scenarios and each scenario was repeated 10 times, with outcomes averaged across the 10 simulations. In this application, it was found that while missingness in excess of 30% has some effect on certain outcomes, there is no evidence to suggest that the missingness mechanism significantly affects results.     

A simulation based method for assessing the statistical significance of logistic regression models after common variable selection procedures

Classification models can demonstrate apparent prediction accuracy even when there is no underlying relationship between the predictors and the response. Variable selection procedures can lead to false positive variable selections and overestimation of true model performance. A simulation study was conducted using logistic regression with forward stepwise, best subsets, and LASSO variable selection methods with varying total sample sizes (20, 50, 100, 200) and numbers of random noise predictor variables (3, 5, 10, 15, 20, 50). Using our critical values can help reduce needless follow-up on variables having no true association with the outcome.     

Optimal and Heuristic Models for Lot Splitting in a Flow Shop

This paper investigates optimal lot-splitting policies in a multiprocess flow shop environment with the objective of minimizing either mean flow time or makespan. Using a quadratic programming approach to the mean flow time problem, we determine the optimal way of splitting a job into smaller sublots under various setup times to run time ratios, number of machines in the flow shop, and number of allowed sublots. Our results come from a deterministic flow shop environment, but also provide insights into the repetitive lots scheme using equal lot splits for job shop scheduling in a stochastic environment. We indicate those conditions in which managers should implement the repetitive lots scheme and where other lot-splitting schemes should work better.     

Shop Performance Implications of Using Cells,Partial Cells,and Remainder Cells*

This paper considers the application of cellular manufacturing (CM) to batch production by exploring the shop floor performance trade‐offs associated with shops employing different levels of CM. The literature has alluded to a continuum that exists between the purely departmentalized job shop and the completely cellular shop. However, the vast majority of CM research exists at the extremes of this continuum. Here, we intend to probe performance relationships by comparing shops that exist at different stages of CM adoption. Specifically, we begin with a hypothetical departmentalized shop found in the CM literature, and in a stepwise fashion, form independent cells. At each stage, flow time and tardiness performance is recorded. Modeling results indicate that, depending on shop conditions and managerial objectives, superior shop performance may be recorded by the job shop, the cell shop, or by one of the shops between these extreme points. In fact, under certain conditions, shops that contain partially formed cells perform better than shops that use completely formed cells. Additional results demonstrate that in order to achieve excellent performance, managers investigating specific layouts need to pay especially close attention to changes in machine utilization as machine groups are partitioned into cells.     

COMPLEX MULTIECHELON INVENTORY SYSTEM MANAGEMENT USING A DYNAMIC SIMULATION MODEL

A dynamic modeling approach to management of multiechelon, multi-indenture inventory systems with repair is addressed. The structure of the model follows that of the U.S. Air Force Reparable Asset Management System. The model is used as a vehicle to discuss the structure of typical multiechelon systems and to illustrate the advantages of a dynamic modeling approach to such systems.     

An Evolutionary Algorithm for Sequencing Production on a Paced Assembly Line

A new sequencing method for mixed‐model assembly lines is developed and tested. This method, called the Evolutionary Production Sequencer (EPS) is designed to maximize production on an assembly line. The performance of EPS is evaluated using three measures: minimum cycle time necessary to achieve 100% completion without rework, percent of items completed without rework for a given cycle time, and sequence “smoothness.” The first two of these measures are based on a simulated production system. Characteristics of the system, such as assembly line station length, assembly time and cycle time, are varied to better gauge the performance of EPS. More fundamental variation is studied by modeling two production systems. In one set of tests, the system consists of an assembly line in isolation (i.e., a single‐level system). In another set of tests, the production system consists of the assembly line and the fabrication system supplying components to the line (i.e., a two‐level system). Sequence smoothness is measured by the mean absolute deviation (MAD) between actual component usage and the ideal usage at each point in the production sequence. The performance of EPS is compared to those of well‐known assembly line sequencing techniques developed by Miltenburg (1989), Okamura and Yamashina (1979), and Yano and Rachamadugu (1991). EPS performed very well under all test conditions when the criterion of success was either minimum cycle time necessary to achieve 100% production without rework or percent of items completed without rework for a given cycle time. When MAD was the criterion of success, EPS was found inferior to the Miltenburg heuristic but better than the other two production‐oriented techniques.     

AN EVALUATION OF SEQUENCING RULES FOR AN ASSEMBLY SHOP

Research relating to sequencing rules in simple job shops has proliferated, but there has not been a corresponding proliferation of research evaluating similar sequencing rules in more complex assembly job shops. In a simple job shop, all operations are performed serially; but an assembly shop encompasses both serial and parallel operations. As a result of the increased complexity of assembly shops, the results associated with the performance of sequencing rules in simple job shops cannot be expected for an assembly shop. In this paper, 11 sequencing rules (some of which are common to simple job shops and some decigned specifically for assembly shops) are evaluated using a simulation analysis of a hypothetical assembly shop. The simulation results are analyzed using an ANOVA procedure that identifies significant differences in the results of several performance measures. Sensitivity analysis also is performed to determine the effect of job structure on the performance of the sequencing rules.     

A Simulation Study of Sequencing Rules in a Kanban-Controlled Flow Shop*

This paper addresses the problem of sequencing in decentralized kanban-controlled flow shops. The kanban production control system considered uses two card types and a constant withdrawal period. The flow shops are decentralized in the sense that sequencing decisions are made at the local workstation level rather than by a centralized scheduling system. Further, there are no material requirements planning (MRP)-generated due dates available to drive dispatching rules such as earliest due date, slack, and critical ratio. Local sequencing rules suitable for the decentralized kanban production-control environment are proposed and tested in a simulation experiment. These rules are designed so that they can be implemented with only the information available at the workstation level. Example sequencing problems are used to show why the shortest processing time rule minimizes neither average work-in-process inventory nor average finished-goods withdrawal kanban waiting time. Further, it is shown how work station supervisors can use the withdrawal period, in addition to the number of kanbans, to manage work-in-process inventories.