An Experimental comparison of linear regression methods used in multi-response design parameter optimization for their estimation and prediction errors

Download

index.pdf

Date

2016

Author

Gökayaz, Gülten

Metadata

Show full item record

Item Usage Stats

211
views

140
downloads

Product and process designers need to find most preferable settings of design parameters to simultaneously achieve multiple quality objectives based on some performance measures such as means and variances of quality characteristics. In these optimization studies, typically empirical models of such performance measures are utilized. These models are usually developed based on data collected through statistically designed experiments using linear regression methods such as Ordinary Least Squares (OLS), Weighted Least Squares (WLS), and Seemingly Unrelated Regression (SUR). In multi-response design parameter optimization (MRDPO) problems, it is assumed that each response has a non-homogeneous variance. Furthermore, responses might be correlated. These linear regression methods might not be appropriate for a particular MRDPO problem due to their restrictive assumptions. Hence, estimation errors associated with model parameters and prediction errors associated with individual observations might be large depending on the problem situation. In this study, we are interested in examining and comparing these errors on a typical MRDPO problem with two responses under different scenarios systematically generated by statistical design of experiments. In addition, we develop a bootstrapping approach to compute joint confidence and prediction regions for estimated mean responses and individual observations, respectively, since these regions are not analytically available for some methods. Our observations based on analysis of experimental results using certain performance measures and graphs of the confidence and prediction regions are presented. Concluding remarks are given and future studies are recommended for generalization of these observations.

Subject Keywords

Regression analysis., Multivariate analysis., Parameter estimation., Stochastic systems.

URI

http://etd.lib.metu.edu.tr/upload/12620193/index.pdf
https://hdl.handle.net/11511/25818

Collections

Graduate School of Natural and Applied Sciences, Thesis

Suggestions

OpenMETU
Core

AN APPROACH FOR DETERMINING THE COST MATRIX OF MULTIVARIATE QUALITY LOSS FUNCTION Dolgun, Leman Esra; Erişkin, Levent; Özkan, Gökçe; Köksal, Gülser (2020-11-01) Multivariate quality loss functions are commonly used in product and process design parameter optimization, which involves simultaneous consideration of multiple responses in determination of the levels of design parameters that provide high quality performance. These functions are also used in statistical tolerancing and quality improvement decision making. This study investigates the bivariate loss function in terms of its ability to represent different values or preferences a decision maker may attribute...
A Case Study on the Utilization of Problem and Solution Domain Measures for Software Size Estimation AYYILDIZ, TÜLİN ERÇELEBİ; Koçyiğit, Altan (2016-09-02) Detailed requirements is the primary input of software size measurement and effort estimation methodologies and a significant amount of time and expertise is needed for size measurement. In order to streamline size measurement and effort estimation, this study exploits the correlations between the problem domain measures such as the number of distinct nouns and distinct verbs in the requirements artifacts and the solution domain measures such as the number of software classes and methods in the correspondin...
An Effort Prediction Model Based on BPM Measures for Process Automation Aysolmaz, Banu; Iren, Deniz; Demirörs, Onur (2013-06-18) BPM software automation projects require different approaches for effort estimation for they are developed based on business process models rather than traditional requirements analysis outputs. In this empirical research we examine the effect of various measures for BPMN compliant business process models on the effort spent to automate those models. Although different measures are suggested in the literature, only a few studies exist that relate these measures to effort estimation. We propose that differen...
An interdisciplinary heuristic evaluation method for universal building design AFACAN, YASEMİN; Erbuğ, Çiğdem (Elsevier BV, 2009-07-01) This study highlights how heuristic evaluation as a usability evaluation method can feed into Current building design practice to conform to universal design principles. It provides a definition of universal usability that is applicable to an architectural design context. It takes the seven universal design principles as a set of heuristics and applies an iterative sequence of heuristic evaluation in a shopping mall, aiming to achieve a cost-effective evaluation process. The evaluation was composed of three...
A semantic discontinuity detection (SDD) method for comparing designers' product expressions with users' product impressions Khalaj, Javad; Pedgley, Owaın Francıs (Elsevier BV, 2019-05-01) Designers exert substantial effort to define an intended user experience (UX) for a product, service or system. However, discontinuities can arise when the actual UX of users differs from the design intent. This paper presents a `Semantic Discontinuity Detection' (SDD) method capable of locating discontinuities between (i) realized product semantics, based on users' initial visual impression of a product, and (ii) intended product semantics, based on the originating designer's visual product expression. The...

Citation Formats

G. Gökayaz, “An Experimental comparison of linear regression methods used in multi-response design parameter optimization for their estimation and prediction errors,” M.S. - Master of Science, Middle East Technical University, 2016.