Advanced input modeling for simulation experimentation

Schmeiser, Bruce W.

doi:10.1145/324138.324173

Cited by 8 publications

(4 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the Continuous method the parameters of a single MVND are estimated by treating all categorical covariates as if they are continuous values, a procedure seen commonly in statistical simulation (12)(13)(14)(15). In order to constrain all biological covariates to be positive, we typically assume a log-normal multivariate distribution.…”

Section: Continuous Methodsmentioning

confidence: 99%

Simulation of Correlated Continuous and Categorical Variables using a Single Multivariate Distribution

Tannenbaum¹,

Holford

Lee

et al. 2006

J Pharmacokinet Pharmacodyn

View full text Add to dashboard Cite

Clinical trial simulations make use of input/output models with covariate effects; the virtual patient population generated for the simulation should therefore display physiologically reasonable covariate distributions. Covariate distribution modeling is one method used to create sets of covariate values (vectors) that characterize individual virtual patients, which should be representative of real subjects participating in clinical trials. Covariates can be continuous (e.g., body weight, age) or categorical (e.g., sex, race). A modeling method commonly used for incorporating both continuous and categorical covariates, the Discrete method, requires the patient population to be divided into subgroups for each unique combination of categorical covariates, with separate multivariate functions for the continuous covariates in each subset. However, when there are multiple categorical covariates this approach can result in subgroups with very few representative patients, and thus, insufficient data to build a model that characterizes these patient groups. To resolve this limitation, an application of a statistical methodology (Continuous method) was conceived to enable sampling of complete covariate vectors, including both continuous and categorical covariates, from a single multivariate function. The Discrete and Continuous methods were compared using both simulated and real data with respect to their ability to generate virtual patient distributions that match a target population. The simulated data sets consisted of one categorical and two correlated continuous covariates. The proportion of patients in each subgroup, correlation between the continuous covariates, and ratio of the means of the continuous covariates in the subgroups were varied. During evaluation, both methods accurately generated the summary statistics and proper proportions of the target population. In general, the Continuous method performed as well as the Discrete method, except when the subgroups, defined by categorical value, had markedly different continuous covariate means, for which, in the authors' experience, there are few clinically relevant examples. The Continuous method allows analysis of the full population instead of multiple subgroups, reducing the number of analyses that must be performed, and thereby increasing efficiency. More importantly, analyzing a larger pool of data increases the precision of the covariance estimates of the covariates, thus improving the accuracy of the description of the covariate distribution in the simulated population.

show abstract

Section: Continuous Methodsmentioning

confidence: 99%

Simulation of Correlated Continuous and Categorical Variables using a Single Multivariate Distribution

Tannenbaum¹,

Holford

Lee

et al. 2006

J Pharmacokinet Pharmacodyn

View full text Add to dashboard Cite

show abstract

“…One of the challenges in the data analysis below is that traditional statistical model selection procedures fail when dealing with extremely large data sets (which we have). Schmeiser (1999) states:…”

Section: Construction Of a Collision Oil Outflow Modelmentioning

confidence: 99%

“…Unfortunately, traditional logistic regression goodness-of-fit tests suffer from the same deficiencies when dealing with extremely large data sets as previously indicated by Schmeiser (1999) (recall (2)). To evaluate visually if ln(y l ), ln(y t ) have explanatory power in terms of the observations z , we generate two sets of residuals r OUT,i = z i − π[ln(y l,i ), ln(y t,i )| β] and r RND,i = z RND,i − π[ln(y l,i ), ln(y t,i )| β], (9) where the index i represents a particular collision scenario from the SR259 report, z i are indicator values (7) describing whether oil outflow occurred in the SR259 collision analysis, Table 5 Logistic regression coefficients for probability of rupture π [ln(y l ), ln(y t )|β] as defined by (8) …”

Section: Data Modeling and Analysis Formentioning

confidence: 99%

An oil outflow model for tanker collisions and groundings

Wiel

Dorp

2009

Ann Oper Res

View full text Add to dashboard Cite

In this paper we have developed an oil outflow model for collision and grounding accidents of tankers. The collision model explicitly links input variables such as tanker hull design (single or double), displacement and speed, striking vessel displacement and speed, and the interaction angle of both vessels to output variables: longitudinal and transversal damage extents of the tanker. Overlaying these damage extents on the tank vessel's design yields an oil outflow volume totaling the capacity of the damaged tank compartments. A similar model is developed for grounding accidents. A total of 80,000 simulation accident scenarios described in the National Research Council SR259 report published in 2001 served as the joint data set of input and output variables used in this "linking" process. The oil outflow model herein was designed keeping computational efficiency in mind to allow for its integration with a maritime transportation system (MTS) simulation. We shall demonstrate the use of the oil outflow model as a final analysis layer to evaluate doublehull effectiveness in a geographic context of an MTS simulation model developed for the oil transportation routes traversing the environmentally sensitive San Juan Islands area in Washington State.

show abstract

“…While point estimates on their own frankly aren't worth much (since you don't know how close or stable or generally good they are), they're a start and can have some properties worth mentioning." This issue is also acknowledged in Schmeiser (1999), where he discusses the differences between Kelton's (1996) stochastic and subjective uncertainty, but which also concludes "the state of the art is far from allowing novice practitioners to build complex input models in the way that they can build complex logical models with today's commercial software".…”

Section: Parameters and Estimatorsmentioning

confidence: 99%

Using worker personality and demographic information to improve system performance prediction

Juran

Schruben

2004

J of Ops Management

View full text Add to dashboard Cite

This paper presents an approach to modeling workers where human performance has a significant impact on system productivity. Highly technical industries such as semiconductor manufacturing and service industries like banking are relying on fewer but more skilled workers. In these systems, productivity depends on worker availability and organization; therefore, modeling system performance may require accurate representations of individual worker behavior. This paper examines the tradeoffs in including information about the demographics and personalities of workers in system performance simulation models. A series of actual and simulated experiments in which personality and demographic data are used in different ways to model the performance of a team of workers is reported. Significant differences are found in predicted system performance demonstrating that model validity depends on the methodology used for modeling workers. These results have practical implication for the managerial processes of recruiting and selecting individual workers, as well as organizing teams of workers and assigning them to tasks.

show abstract

Advanced input modeling for simulation experimentation

Cited by 8 publications

References 59 publications

Simulation of Correlated Continuous and Categorical Variables using a Single Multivariate Distribution

Simulation of Correlated Continuous and Categorical Variables using a Single Multivariate Distribution

An oil outflow model for tanker collisions and groundings

Using worker personality and demographic information to improve system performance prediction

Contact Info

Product

Resources

About