skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: A non‐approximate method for generating G$G$‐optimal RSM designs
We present a nonapproximate computational method for generating ‐optimal designs in response surface methodology (RSM) settings usingGloptipoly, a global polynomial optimizer. Traditional approaches use a grid approximation for computing a candidate design's ‐score.Gloptipolycan find the global optimum of high‐order polynomials thus making it suitable for computing a design's ‐score, that is, its maximum scaled prediction variance, which, for second‐order models, is a quartic polynomial function of the experimental factors. We demonstrate the efficacy and performance of our method through comprehensive application to well‐published examples, and illustrate, for the first time, its application to generating ‐optimal designs supporting models of order greater than 2. This work represents the first non‐approximate computational approach to solving the ‐optimal design problem. This advancement opens new possibilities for finding ‐optimal designs beyond second‐order RSM models.  more » « less
Award ID(s):
2412020 2134409
PAR ID:
10552229
Author(s) / Creator(s):
; ;
Publisher / Repository:
Quality and Reliability Engineering, International
Date Published:
Journal Name:
Quality and Reliability Engineering International
ISSN:
0748-8017
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; (, Networks)
    Abstract The concept of branch decomposition was first introduced by Robertson and Seymour in their proof of the Graph Minors Theorem, and can be seen as a measure of the global connectivity of a graph. Since then, branch decomposition and branchwidth have been used for computationally solving combinatorial optimization problems modeled on graphs and matroids. General branchwidth is the extension of branchwidth to any symmetric submodular function defined over a finite set. General branchwidth encompasses graphic branchwidth, matroidal branchwidth, and rankwidth. A tangle basis is related to a tangle, a notion also introduced by Robertson and Seymour; however, a tangle basis is more constructive in nature. It was shown in [I. V. Hicks. Graphs, branchwidth, and tangles! Oh my!Networks, 45:55‐60, 2005] that a tangle basis of orderkis coextensive to a tangle of orderk. In this paper, we revisit the construction of tangle bases computationally for other branchwidth parameters and show that the tangle basis approach is still competitive for computing optimal branch decompositions for general branchwidth. 
    more » « less
  2. ; ; ; (, ACM Transactions on Embedded Computing Systems)
    Artificial intelligence (AI) provides versatile capabilities in applications such as image classification and voice recognition that are most useful in edge or mobile computing settings. Shrinking these sophisticated algorithms into small form factors with minimal computing resources and power budgets requires innovation at several layers of abstraction: software, algorithmic, architectural, circuit, and device-level innovations. However, improvements to system efficiency may impact robustness and vice-versa. Therefore, a co-design framework is often necessary to customize a system for its given application. A system that prioritizes efficiency might use circuit-level innovations that introduce process variations or signal noise into the system, which may use software-level redundancy in order to compensate. In this tutorial, we will first examine various methods of improving efficiency and robustness in edge AI and their tradeoffs at each level of abstraction.Then, we will outline co-design techniques for designing efficient and robust edge AI systems, using federated learning as a specific example to illustrate the effectiveness of co-design. 
    more » « less
  3. ; ; ; (, Geophysical Journal International)
    SUMMARY This paper revisits and extends the adjoint theory for glacial isostatic adjustment (GIA) of Crawford et al. (2018). Rotational feedbacks are now incorporated, and the application of the second-order adjoint method is described for the first time. The first-order adjoint method provides an efficient means for computing sensitivity kernels for a chosen objective functional, while the second-order adjoint method provides second-derivative information in the form of Hessian kernels. These latter kernels are required by efficient Newton-type optimization schemes and within methods for quantifying uncertainty for non-linear inverse problems. Most importantly, the entire theory has been reformulated so as to simplify its implementation by others within the GIA community. In particular, the rate-formulation for the GIA forward problem introduced by Crawford et al. (2018) has been replaced with the conventional equations for modelling GIA in laterally heterogeneous earth models. The implementation of the first- and second-order adjoint problems should be relatively easy within both existing and new GIA codes, with only the inclusions of more general force terms being required. 
    more » « less
  4. ; (, Statistical Analysis and Data Mining: The ASA Data Science Journal)
    Abstract In this work, we develop a method namedTwinningfor partitioning a dataset into statistically similar twin sets.Twinningis based onSPlit, a recently proposed model‐independent method for optimally splitting a dataset into training and testing sets.Twinningis orders of magnitude faster than theSPlitalgorithm, which makes it applicable to Big Data problems such as data compression.Twinningcan also be used for generating multiple splits of a given dataset to aid divide‐and‐conquer procedures andk‐fold cross validation. 
    more » « less
  5. ; ; (, Numerical Methods for Partial Differential Equations)
    Abstract We begin with a treatment of the Caputo time‐fractional diffusion equation, by using the Laplace transform, to obtain a Volterra integro‐differential equation. We derive and utilize a numerical scheme that is derived in parallel to the L1‐method for the time variable and a standard fourth‐order approximation in the spatial variable. The main method derived in this article has a rate of convergence ofO(kα + h4)foru(x,t) ∈ Cα([0,T];C6(Ω)),0 < α < 1, which improves previous regularity assumptions that requireC2[0,T]regularity in the time variable. We also present a novel alternative method for a first‐order approximation in time, under a regularity assumption ofu(x,t) ∈ C1([0,T];C6(Ω)), while exhibiting order of convergence slightly more thanO(k)in time. This allows for a much wider class of functions to be analyzed which was previously not possible under the L1‐method. We present numerical examples demonstrating these results and discuss future improvements and implications by using these techniques. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email