A5008353745- Matrix Theory and Algorithms
- Efficiency Analysis Using DEA
- Advanced Optimization Algorithms Research
- Water resources management and optimization
- Sparse and Compressive Sensing Techniques
- Advanced Neural Network Applications
- Economic and Environmental Valuation
- Economic and Technological Innovation
- Optimization and Mathematical Programming
- Engineering Applied Research
- Iterative Methods for Nonlinear Equations
- Machine Learning and Data Classification
- Advanced Multi-Objective Optimization Algorithms
- Electric Power System Optimization
- Multi-Criteria Decision Making
- Stochastic Gradient Optimization Techniques
University of Minnesota
2021-2024
Twin Cities Orthopedics
2021
Lawrence Berkeley National Laboratory
2021
Rose–Hulman Institute of Technology
2019
Classifications organize entities into categories that identify similarities within a category and discern dissimilarities among categories, they powerfully classify information in support of analysis. We propose new classification scheme premised on the reality imperfect data. Our computational model uses uncertain data envelopment analysis to define classification's proximity equitable efficiency, which is an aggregate measure intra-similarity categories. process has two overriding...
We present a new software, HYPPO, that enables the automatic tuning of hyperparameters various deep learning (DL) models. Unlike other hyperparameter optimization (HPO) methods, HYPPO uses adaptive surrogate models and directly accounts for uncertainty in model predictions to find accurate reliable make robust predictions. Using asynchronous nested parallelism, we are able significantly alleviate computational burden training complex architectures quantifying uncertainty. is implemented...
In this paper, we introduce a new optimization algorithm that is well suited for solving parameter estimation problems. We call our method cubic regularized Newton with affine scaling (CRNAS). contrast to so-called first-order methods which rely solely on the gradient of objective function, utilizes Hessian objective. As result it able focus points satisfying second-order optimality conditions, as opposed simply converge critical points. This an important feature in problems where function...
This paper presents and analyzes the first matrix optimization model which allows general coordinate spectral constraints. The breadth of problems our covers is exemplified by a lengthy list examples from literature, including semidefinite programming, completion, quadratically constrained quadratic programs (QCQPs), we demonstrate enables completely novel formulations numerous problems. Our solution methodology leverages factorization manifold to develop an equivalent reformulation for...
We investigate how to solve smooth matrix optimization problems with general linear inequality constraints on the eigenvalues of a symmetric matrix. present solution methods obtain exact global minima for objective functions, i.e., $F(X) = \langle C, X \rangle$, and perform projections onto eigenvalue constraint set. Two first-order algorithms are developed stationary points non-convex functions. Both proven converge sublinearly when set is convex. Numerical experiments demonstrate...
In this article we use the Unified Transform Method to study boundary value problems for a hyperbolic system of partial differential equations from relativistic quantum mechanics.Specifically, derive solutions Dirac equation in both massive and massless cases on half-line finite interval using method.
Classifications organize entities into categories that identify similarities within a category and discern dissimilarities among categories, they powerfully classify information in support of analysis. We propose new classification scheme premised on the reality imperfect data. Our computational model uses uncertain data envelopment analysis to define classification's proximity equitable efficiency, which is an aggregate measure intra-similarity categories. process has two overriding...
Gerrymandering voting districts is one of the most salient concerns contemporary American society, and creation new maps, along with their subsequent legal challenges, speaks for much our modern political discourse. The legal, societal, debate over serviceable demands a concept fairness, which loosely characterized, but amorphous, that has evaded precise definition. We advance paradigm to compare maps avoids pitfalls associated an priori metric being used uniformly assess maps. Our...