Matthew Collette
University of Michigan
59 Papers
351 Citations
Matthew Collette is an academic researcher from University of Michigan. The author has contributed to research in topics: Computer science & Bayesian network. The author has an hindex of 13, co-authored 54 publications. Previous affiliations of Matthew Collette include Science Applications International Corporation & University of Newcastle.
Chat about Author
Papers
Improving surrogate-assisted variable fidelity multi-objective optimization using a clustering algorithm
Yan Liu,Matthew Collette +1 more
- 01 Nov 2014
TL;DR: In this article, the authors extended variable fidelity optimization framework to include multiple surrogates and used a k-means clustering algorithm to partition model data into local surrogate models to solve the large sample size surrogate-modeling problem.
49
•Dissertation
The strength and reliability of aluminium stiffened panels
Matthew Collette
- 01 Jan 2005
TL;DR: In this article, a reliability-based hot-spot S-N fatigue prediction method is developed for welded connections, including an analysis of the material and prediction uncertainty values and a comparison with existing design codes.
44
Ultimate compressive strength design methods of aluminum welded stiffened panel structures for aerospace, marine and land-based applications: A benchmark study
TL;DR: In this article, the authors compared some useful ULS methods adopted for the design of aerospace, marine and land-based aluminum structures, and discussed a common practice for aerospace and marine and civil engineering welded stiffened panel applications.
43
A multi-objective variable-fidelity optimization method for genetic algorithms
TL;DR: In this article, a variable-fidelity optimization (VFO) scheme for multi-objective genetic algorithms is presented, which uses a low and high fidelity version of the objective function with a Kriging scaling model to interpolate between them.
33
Efficient optimization of reliability-constrained structural design problems including interval uncertainty
TL;DR: In this article, an interval uncertainty formulation for exploring the impact of epistemic uncertainty on reliability-constrained design performance is proposed, where an adaptive surrogate modeling framework is developed to locate the lowest reliability value within a multi-dimensional interval.
30