Alan M. Polansky
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Alan M. Polansky has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
Robert L. Mason
Southwest Research Institute, San Antonio, TX 78228-0510
1 shared publicationWilliam. R. Schucany
Southern Methodist University , Dallas , USA
1 shared publicationYoun-Min Chou
The University of Texas at San Antonio, San Antonio, TX 78249-0664
1 shared publication
Latest Publications
- 1 April 2014
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 46 (2) , 150-170
Assessing the Capability of a Manufacturing Process Using Nonparametric Bayesian Density Estimation
The capability of a manufacturing process refers to the ability of the process to produce items that are within set specifications. The determination of the capability of a stable manufacturing process with a multivariate quality characteristic using standard methods usually requires the assumption that the quality characteristic of interest follows a multivariate normal distribution, an assumption that is difficult to assess in practice. Departures from this assumption can result in erroneous conclusions, which can be costly to the manufacturer. In this paper, we propose assessing the capability of a process using a nonparametric Bayesian framework. This framework relies on using mixtures of Dirichlet processes to elicit a prior on the multivariate distribution of the quality characteristic. The posterior distribution of the distribution of the quality characteristic can then be used to induce a posterior distribution on the process fallout rate, the mean distance between the quality characteristic and a specified target value, or on an associated process capability index. The methodology is demonstrated through three examples using real-world data. Particular emphasis in these examples is given toward specifying the parameter values required to specify the prior distribution and on interpretation of the results.
- 1 April 2014
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 46 (2) , 150-170
Assessing the Capability of a Manufacturing Process Using Nonparametric Bayesian Density Estimation
The capability of a manufacturing process refers to the ability of the process to produce items that are within set specifications. The determination of the capability of a stable manufacturing process with a multivariate quality characteristic using standard methods usually requires the assumption that the quality characteristic of interest follows a multivariate normal distribution, an assumption that is difficult to assess in practice. Departures from this assumption can result in erroneous conclusions, which can be costly to the manufacturer. In this paper, we propose assessing the capability of a process using a nonparametric Bayesian framework. This framework relies on using mixtures of Dirichlet processes to elicit a prior on the multivariate distribution of the quality characteristic. The posterior distribution of the distribution of the quality characteristic can then be used to induce a posterior distribution on the process fallout rate, the mean distance between the quality characteristic and a specified target value, or on an associated process capability index. The methodology is demonstrated through three examples using real-world data. Particular emphasis in these examples is given toward specifying the parameter values required to specify the prior distribution and on interpretation of the results.
- 1 January 2010
- Published by Elsevier in Computational Statistics & Data Analysis
- Vol. 54 (1) , 233-244
Ordered inference using observed confidence levels
- 1 January 2010
- Published by Elsevier in Computational Statistics & Data Analysis
- Vol. 54 (1) , 233-244
Ordered inference using observed confidence levels
- 1 July 2006
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 38 (3) , 254-266
Permutation Methods for Comparing Process Capabilities
The statistical comparison of competing manufacturing processes is an important aspect of statistical quality control that aids quality managers in the choice of potential suppliers of products, manufacturing methodologies, or proposed adjustments to the manufacturing process. The selection criteria, in the absence of considerations such as cost, is based on a quality metric, such as the capability of each manufacturing process. Because quality metrics are estimated based on sample process data, the inherent variability in the estimates must be accounted for when selecting the best manufacturing process. In this paper, we consider solutions to this problem based on permutation testing methodology. In the case of two processes, the methodology is based on a simple permutation test of the null hypothesis that the two processes have equal capability. In the case of more than two processes, multiple-comparison techniques are used in conjunction with the proposed permutation tests. The advantage of using the permutation methods is that the significance levels of the permutation tests are exact regardless of the distribution of the process data. The methodology is demonstrated using several examples, and the potential performance of the methods are investigated empirically.
- 1 July 2006
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 38 (3) , 254-266
Permutation Methods for Comparing Process Capabilities
The statistical comparison of competing manufacturing processes is an important aspect of statistical quality control that aids quality managers in the choice of potential suppliers of products, manufacturing methodologies, or proposed adjustments to the manufacturing process. The selection criteria, in the absence of considerations such as cost, is based on a quality metric, such as the capability of each manufacturing process. Because quality metrics are estimated based on sample process data, the inherent variability in the estimates must be accounted for when selecting the best manufacturing process. In this paper, we consider solutions to this problem based on permutation testing methodology. In the case of two processes, the methodology is based on a simple permutation test of the null hypothesis that the two processes have equal capability. In the case of more than two processes, multiple-comparison techniques are used in conjunction with the proposed permutation tests. The advantage of using the permutation methods is that the significance levels of the permutation tests are exact regardless of the distribution of the process data. The methodology is demonstrated using several examples, and the potential performance of the methods are investigated empirically.
- 10 March 2005
- Published by Wiley in Quality and Reliability Engineering International
- Vol. 21 (6) , 633-653
A General Framework for Constructing Control Charts
A general framework for the construction of control charts is presented. The method is based on using the density of the sample subgroup statistic as a measure of how unusual newly observed subgroups are. This methodology includes, as special cases, many common control chart techniques. The method is also easily applied to multivariate and multimodal situations. A non‐parametric control chart is implemented by estimating the density of the sample subgroup statistic using a kernel estimator of the bootstrap distribution of the observed subgroup statistics. Several examples of the method are presented in the parametric and non‐parametric situations. The potential performance of the non‐parametric version of the method is demonstrated through an empirical study of average run length properties. Copyright © 2005 John Wiley & Sons, Ltd.
- 10 March 2005
- Published by Wiley in Quality and Reliability Engineering International
- Vol. 21 (6) , 633-653
A General Framework for Constructing Control Charts
A general framework for the construction of control charts is presented. The method is based on using the density of the sample subgroup statistic as a measure of how unusual newly observed subgroups are. This methodology includes, as special cases, many common control chart techniques. The method is also easily applied to multivariate and multimodal situations. A non‐parametric control chart is implemented by estimating the density of the sample subgroup statistic using a kernel estimator of the bootstrap distribution of the observed subgroup statistics. Several examples of the method are presented in the parametric and non‐parametric situations. The potential performance of the non‐parametric version of the method is demonstrated through an empirical study of average run length properties. Copyright © 2005 John Wiley & Sons, Ltd.
- 29 October 2003
- Published by Wiley in Statistics in Medicine
- Vol. 22 (22) , 3461-3471
Selecting the best treatment in designed experiments
- 29 October 2003
- Published by Wiley in Statistics in Medicine
- Vol. 22 (22) , 3461-3471
Selecting the best treatment in designed experiments
- 1 September 2000
- Published by Wiley in Canadian Journal of Statistics
- Vol. 28 (3) , 501-516
Stabilizing bootstrap‐tconfidence intervals for small samples
A major use of the bootstrap methodology is in the construction of nonparametric confidence intervals. Although no consensus has yet been reached on the best way to proceed, theoretical and empirical evidence indicate that bootstra.‐tintervals provide a reasonable solution to this problem. However, when applied to small data sets, these intervals can be unusually wide and unstable. The author presents techniques for stabilizing bootstra.‐tintervals for small samples. His methods are motivated theoretically and investigated though simulations.
- 1 September 2000
- Published by Wiley in Canadian Journal of Statistics
- Vol. 28 (3) , 501-516
Stabilizing bootstrap‐tconfidence intervals for small samples
A major use of the bootstrap methodology is in the construction of nonparametric confidence intervals. Although no consensus has yet been reached on the best way to proceed, theoretical and empirical evidence indicate that bootstra.‐tintervals provide a reasonable solution to this problem. However, when applied to small data sets, these intervals can be unusually wide and unstable. The author presents techniques for stabilizing bootstra.‐tintervals for small samples. His methods are motivated theoretically and investigated though simulations.
- 1 July 2000
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 32 (3) , 284-289
An Algorithm for Computing a Smooth Nonparametric Process Capability Estimate
A program written in ANSI C that computes a kernel based estimate of the proportion nonconforming is presented. The bandwidth for the kernel estimate is computed using several multistage plug-in methods that have been shown to perform well in practice. These bandwidths, along with estimates of the process fallout rate, are provided by the program.
- 1 July 2000
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 32 (3) , 284-289
An Algorithm for Computing a Smooth Nonparametric Process Capability Estimate
A program written in ANSI C that computes a kernel based estimate of the proportion nonconforming is presented. The bandwidth for the kernel estimate is computed using several multistage plug-in methods that have been shown to perform well in practice. These bandwidths, along with estimates of the process fallout rate, are provided by the program.
- 1 July 1999
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 31 (3) , 345-350
An Algorithm for Fitting Johnson Transformations to Non-Normal Data
A FORTRAN program which selects the estimated best-fit Johnson transformation for a set of non-normal data is presented. The parameter estimates are computed using sample percentiles, and the quality of the fit is assessed using the Shapiro-Wilk test statistic on the transformed data. The estimated transformation with the largest p-value is provided along with the transformed data.
- 1 July 1999
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 31 (3) , 345-350
An Algorithm for Fitting Johnson Transformations to Non-Normal Data
A FORTRAN program which selects the estimated best-fit Johnson transformation for a set of non-normal data is presented. The parameter estimates are computed using sample percentiles, and the quality of the fit is assessed using the Shapiro-Wilk test statistic on the transformed data. The estimated transformation with the largest p-value is provided along with the transformed data.
- 1 January 1998
- Published by Wiley in Quality and Reliability Engineering International
- Vol. 14 (1) , 43-48
A smooth nonparametric approach to process capability
- 1 January 1998
- Published by Wiley in Quality and Reliability Engineering International
- Vol. 14 (1) , 43-48
A smooth nonparametric approach to process capability
- 1 November 1997
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 59 (4) , 821-838
Kernel Smoothing to Improve Bootstrap Confidence Intervals
- 1 November 1997
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 59 (4) , 821-838
Kernel Smoothing to Improve Bootstrap Confidence Intervals