# Embedded Ridge Approximations: Constructing Ridge Approximations Over Localized Scalar Fields For Improved Simulation-Centric Dimension Reduction

@article{Wong2019EmbeddedRA, title={Embedded Ridge Approximations: Constructing Ridge Approximations Over Localized Scalar Fields For Improved Simulation-Centric Dimension Reduction}, author={Chun Yui Wong and Pranay Seshadri and Geoffrey T. Parks and Mark A. Girolami}, journal={ArXiv}, year={2019}, volume={abs/1907.07037} }

Many quantities of interest (qois) arising from differential-equation-centric models can be resolved into functions of scalar fields. Examples of such qois include the lift over an airfoil or the displacement of a loaded structure; examples of corresponding fields are the static pressure field in a computational fluid dynamics solution, and the strain field in the finite element elasticity analysis. These scalar fields are evaluated at each node within a discretized computational domain. In…

## One Citation

### Optimization by moving ridge functions: derivative-free optimization for computationally intensive functions

- Computer ScienceEngineering Optimization
- 2021

A novel derivative-free algorithm, called optimization by moving ridge functions (OMoRF), for unconstrained and bound-constrained optimization is presented. This algorithm couples trust region…

## 33 References

### A Lanczos-Stieltjes method for one-dimensional ridge function approximation and integration

- Computer Science
- 2018

The key elements of the method are approximating the univariate density of the given linear combination of inputs by repeated convolutions and a Lanczos-Stieltjes method for constructing orthogonal polynomials and Gaussian quadrature.

### Erratum: Active Subspace Methods in Theory and Practice: Applications to Kriging Surfaces

- Computer ScienceSIAM J. Sci. Comput.
- 2014

This work presents a method to first detect the directions of the strongest variability using evaluations of the gradient and subsequently exploit these directions to construct a response surface on a low-dimensional subspace---i.e., the active subspace ---of the inputs.

### Data-Driven Polynomial Ridge Approximation Using Variable Projection

- Computer ScienceSIAM J. Sci. Comput.
- 2018

The Gauss-Newton method has superior theoretical guarantees and faster convergence than the alternating approach for polynomial ridge approximation earlier proposed by Constantine, Eftekhari, Hokanson, and Ward.

### Dimension Reduction via Gaussian Ridge Functions

- Computer Science, MathematicsSIAM/ASA J. Uncertain. Quantification
- 2019

This paper develops an iterative algorithm that optimizes over the Grassmann manifold to compute the subspace, followed by an optimization of the hyperparameters of the Gaussian process to approximate a Gaussian ridge function.

### Multifidelity Dimension Reduction via Active Subspaces

- Computer ScienceSIAM J. Sci. Comput.
- 2020

This work provides a non-asymptotic analysis of the number of gradient evaluations sufficient to achieve a prescribed error in the active subspace matrix, both in expectation and with high probability.

### Uncertainty quantification of the transonic flow around the RAE 2822 airfoil

- Computer Science
- 2010

Motivated by the RAE 2822 test case, a Simplex Elements Stochastic Collocation (SESC) method is developed that combines a robust approximation of discon-tinuous responses with an efficient discretization in multi–dimensional probability spaces.

### Inverse regression for ridge recovery: a data-driven approach for parameter reduction in computer experiments

- Mathematics, Computer ScienceStat. Comput.
- 2020

This work interprets SDR methods sliced inverse regression (SIR) and sliced average variance estimation (SAVE) as estimating the directions of a ridge function, which is a composition of a low-dimensional linear transformation with a nonlinear function.

### Computing active subspaces efficiently with gradient sketching

- Computer Science2015 IEEE 6th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP)
- 2015

This work numerically study two methods for estimating the necessary eigenpairs of a matrix derived from the function's gradient using only linear measurements of the function’s gradient.

### Gradient-Based Dimension Reduction of Multivariate Vector-Valued Functions

- Mathematics, Computer ScienceSIAM J. Sci. Comput.
- 2020

A gradient-based method for detecting and exploiting low-dimensional input parameter dependence of multivariate functions and reveals that the choice of norm on the codomain of the function can have a significant impact on the function's low- dimensional approximation.

### Leakage Uncertainties in Compressors: The Case of Rotor 37

- Engineering, Physics
- 2015

This paper revisits an old problem of validating computational fluid dynamics simulations with experiments in turbomachinery. The case considered here is NASA rotor 37. Prior computational fluid…