Computational Mathematics

Adaptive Finite Element Methods and A Posteriori Error Estimation

Development of adaptive mesh refinement strategies and rigorous error bounds for finite element discretizations of partial differential equations.

RESEARCH GAP FRONTIERS

Machine Learning Enhanced Numerical Solvers

Integration of neural networks and deep learning with classical numerical methods to accelerate convergence and improve solution accuracy.

RESEARCH GAP FRONTIERS

High-Dimensional Uncertainty Quantification and Sampling

Advanced techniques for propagating uncertainty through complex mathematical models using sparse grids and quasi-Monte Carlo methods.

RESEARCH GAP FRONTIERS

Structure-Preserving Integrators for Hamiltonian Systems

Development of symplectic and energy-stable time integration schemes that maintain geometric properties of Hamiltonian dynamics.

RESEARCH GAP FRONTIERS

Multiscale Modeling and Homogenization Techniques

Computational methods for bridging microscopic and macroscopic scales in heterogeneous materials and complex systems.

RESEARCH GAP FRONTIERS

Rational Approximation and Krylov Subspace Methods

Construction of rational functions and Krylov methods for solving large-scale linear systems and eigenvalue problems.

RESEARCH GAP FRONTIERS

Discontinuous Galerkin Methods for Hyperbolic Conservation Laws

High-order discretization schemes for capturing shock discontinuities and preserving physical constraints in nonlinear hyperbolic equations.

RESEARCH GAP FRONTIERS

Physics-Informed Neural Networks for Differential Equations

Mesh-free surrogate models that encode physical laws as constraints within neural network training for solving differential equations.

RESEARCH GAP FRONTIERS

Spectral Methods for Variable Coefficient Problems

Development of high-accuracy spectral discretizations for differential equations with spatially varying coefficients and irregular domains.

Domain Decomposition Methods for Large-Scale Computing

Parallel algorithms that partition computational domains to enable efficient distributed solving of multiphysics and coupled problems.

Inverse Problems and Regularization Theory

Mathematical frameworks for recovering unknown parameters and initial conditions from noisy observational data in ill-posed problems.

Fast Algorithms for Nonlocal and Fractional Operators

Efficient computational schemes for handling integral and fractional derivative operators appearing in anomalous diffusion and peridynamics.

Topology Optimization and Shape Calculus

Computational methods for optimizing material distribution and domain shapes to achieve desired mechanical or physical performance.

Monte Carlo Methods and Variance Reduction Techniques

Advanced probabilistic simulation techniques including multilevel Monte Carlo and importance sampling for uncertainty quantification.

Stochastic Differential Equations and Path Sampling

Numerical integration methods for stochastic systems with applications to molecular dynamics and Bayesian inference.

Finite Difference Schemes for Nonlinear Waves

Accurate and stable discretizations for soliton equations and nonlinear wave phenomena with preservation of conservation laws.

Computational Algebraic Geometry and Polynomial Systems

Numerical methods for solving systems of multivariate polynomial equations using Gröbner bases and homotopy continuation.

Multigrid Algorithms and Fast Solvers

Development of multilevel iterative methods achieving near-optimal computational complexity for discretized elliptic problems.

Reduced-Order Modeling via Proper Orthogonal Decomposition

Construction of low-dimensional surrogates from high-dimensional simulation data for parametric studies and real-time applications.

Boundary Integral Methods and Fast Multipole Algorithms

Efficient computation of solutions to boundary value problems via Green''s functions and hierarchical acceleration techniques.

Numerical Methods for Matrix Equations and Tensor Problems

Algorithms for solving Sylvester, Lyapunov, and Riccati equations with applications to high-order tensor decompositions.

Conservative Schemes for Nonlinear Schroedinger Equations

Structure-preserving discretizations maintaining mass and energy conservation for quantum mechanical and optical wave problems.

Iterative Methods for Saddle Point Systems

Preconditioned Krylov subspace solvers for mixed finite element formulations of incompressible flow and constraint problems.

Fluid-Structure Interaction Simulations

Coupling algorithms for solving strongly interacting fluid dynamics and solid mechanics with monolithic or partitioned approaches.

Composite Finite Element Methods

Numerical schemes combining multiple discretization strategies for multiphase flows and heterogeneous material systems.

Optimization Under Uncertainty and Robust Design

Algorithms for finding optimal designs that remain feasible across uncertain parameter ranges using stochastic programming.

Isogeometric Analysis and NURBS Discretizations

Integration of computer-aided design and finite element analysis using non-uniform rational B-splines for exact geometry representation.

Virtual Element Methods and Polytopal Meshes

Advanced discretization schemes handling arbitrary polygonal and polyhedral elements with applicability to complex geometries.

Singular Perturbation Problems and Asymptotic Analysis

Numerical treatment of equations with boundary layers and internal transitions using singular perturbation and WKB methods.

Immersed Boundary Methods for Moving Interfaces

Algorithms for capturing dynamics of flexible structures immersed in fluids without explicit boundary tracking.

Fast Wavelet Transforms and Multiresolutional Analysis

Development of orthogonal wavelet bases and algorithms for efficient compression and solution of multiscale problems.

Discontinuity Capturing and Total Variation Diminishing Schemes

Nonlinear numerical methods that detect and resolve shocks and steep gradients without spurious oscillations.

Quantum Computing for Differential Equations

Variational quantum algorithms and hybrid classical-quantum approaches for solving partial differential equations on quantum hardware.

Preconditioners for Ill-Conditioned Systems

Construction of matrix preconditioners improving iterative solver convergence for highly ill-conditioned discretized equations.

Level Set Methods and Interface Tracking

Implicit representation of moving interfaces and boundaries using signed distance functions for multiphase flow problems.

Petrov-Galerkin Methods and Stabilization Techniques

Asymmetric Galerkin methods with stabilization operators for advection-dominated transport and incompressible flow simulation.

Constraint Handling in Optimization Algorithms

Methods for incorporating equality and inequality constraints in constrained optimization including penalty and augmented Lagrangian approaches.

Backward Error Analysis and Numerical Stability

Rigorous frameworks for analyzing how computational errors propagate through algorithms and determining algorithm stability properties.

Generalized Eigenvalue Problems and Spectral Analysis

Algorithms for computing eigenvalues and eigenvectors of matrix pencils arising in vibration and stability analysis.

Coupling of Heterogeneous Spatial and Temporal Discretizations

Methods for solving multidomain and multiscale problems using different discretization types in different regions.

Numerical Integration of Singular Integrands

Specialized quadrature rules for approximating integrals with singularities using transformation techniques and weighted formulas.

Lattice Boltzmann Methods for Fluid Dynamics

Kinetic-based mesoscopic simulation method for fluid flow combining simplicity with ability to handle complex boundaries and multiphase flows.

Gradient Flows and Variational Time Stepping

Energy-stable time integration schemes based on gradient flow structures for dissipative systems and chemical reactions.

Compressive Sensing and Sparse Recovery

Algorithms for reconstructing high-dimensional data from undersampled measurements exploiting sparsity in some basis.

Nonlinear Preconditioning Strategies

Application of physics-based and nonlinear transformations as preconditioners for improving convergence of nonlinear iterations.

Extrapolation Methods and Richardson Acceleration

Sequence acceleration techniques combining approximations at different discretization levels to improve convergence rates.

Partition of Unity Methods and Meshfree Schemes

Discretization methods requiring only point clouds without mesh generation using smooth weight functions and local approximations.

Weighted Essentially Non-Oscillatory Reconstruction Schemes

High-order finite volume methods using adaptive weighted reconstruction for accurate capture of smooth and discontinuous features.

Continuous and Discrete Adjoint Methods

Efficient computation of gradients and sensitivity derivatives with respect to design parameters for optimization and UQ.

Modular Operator Splitting for Coupled Systems

Decomposition strategies for solving coupled multiphysics problems by independently treating subsystems with appropriate interface conditions.

Exponential Integrators for Stiff Equations

Development of high-order time integration schemes using matrix exponentials and Krylov approximations for efficiently solving stiff differential equations.

Surrogate Modeling and Gaussian Process Regression

Construction of computationally efficient surrogate models using Gaussian processes and Bayesian optimization for expensive simulations and design problems.

Structure-Preserving Numerical Methods for Dissipative Systems

Design of discretization schemes that maintain energy dissipation and thermodynamic consistency for gradient flows and non-equilibrium systems.

Numerical Methods for Fractional Calculus Problems

Development of accurate and efficient algorithms for solving fractional differential equations with Caputo and Riemann-Liouville derivatives.

Mimetic Finite Difference Methods and Discretizations

Construction of discrete operators that preserve fundamental properties of vector calculus and conservation laws in mixed-media applications.

Localized Orthogonal Decomposition Techniques

Development of efficient coarse-grained approximations for multiscale problems through localized orthogonal decomposition and basis construction.

Computation of Lyapunov Exponents and Chaotic Systems

Numerical algorithms for accurately computing Lyapunov exponents and analyzing stability and predictability in nonlinear dynamical systems.

Optimization Algorithms for Machine Learning Applications

Analysis and development of stochastic gradient methods, adaptive learning rates, and distributed optimization algorithms for deep learning.

Computational Methods for Optimal Control Problems

Numerical techniques including direct and indirect methods for solving constrained optimal control problems with applications to aerospace and robotics.

Spectral Element Methods and Gauss-Lobatto Quadratures

High-order finite element methods combining spectral accuracy with geometric flexibility using Gauss-Lobatto nodal distributions.

Subspace Iteration and Generalized Davidson Methods

Development of iterative eigensolvers for computing interior eigenvalues and exploring low-rank approximations in large-scale eigenvalue problems.

Numerical Methods for Mean Field Games

Computational techniques for solving coupled forward-backward systems arising from mean field game theory and multi-agent optimization.

Time Integration for Advection-Dominated Equations

Design of stable and accurate explicit and implicit time stepping methods that handle large Courant numbers and transport phenomena.

Finite Element Methods on Curved Surfaces

Development of surface FEM discretizations for PDEs defined on manifolds including error analysis and geometric approximation.

Hybrid High-Order Methods for PDEs

Construction of hybrid discretization schemes combining cell and face unknowns for improved approximation on polyhedral meshes.

Numerical Homogenization and Effective Media Theory

Computational methods for deriving effective material properties from heterogeneous microstructures and multiscale wave propagation.

Continuation Methods and Bifurcation Analysis

Numerical techniques for tracking solution branches, detecting bifurcations, and analyzing stability in parameterized nonlinear systems.

Radial Basis Function Approximation Theory

Theory and algorithms for RBF-based meshfree approximations with applications to scattered data interpolation and PDE solving.

Numerical Linear Algebra for Tensor Computations

Algorithms for tensor decomposition, Tucker formats, tensor train representations, and efficient tensor-vector operations.

Coarse Graining and Renormalization Methods

Development of systematic approaches for deriving reduced models from molecular dynamics and microscopic simulations via renormalization.

Numerical Methods for Nonlocal Conservation Laws

Discretization techniques for conservation laws with nonlocal fluxes and integral operators appearing in peridynamics and anomalous transport.

Accurate Computation of Matrix Functions

Algorithms for computing matrix exponentials, logarithms, and other functions with high accuracy and applications to solving matrix equations.

Numerical Methods for Kinetic Boltzmann Equations

Computational schemes for high-dimensional kinetic equations including spectral methods and asymptotic-preserving integrators.

Parameterized Model Reduction and Basis Methods

Development of parametric reduced-order models using greedy algorithms and reduced basis methods for parametric PDE solving.

Numerical Methods for Coupled Bulk-Surface PDEs

Discretization schemes for coupled systems involving bulk and surface equations with applications to active transport and cell biology.

Convex Relaxation and Semidefinite Programming

Convex relaxation techniques and interior-point methods for solving large-scale optimization and approximation problems efficiently.

Numerical Methods for McKean-Vlasov Equations

Computational algorithms for mean-field particle systems and their continuum limits with applications to sampling and inference.

Higher-Order Weak Approximation for SDEs

Development of high-order weak approximation schemes for SDEs achieving accuracy in expectations with lower computational cost than strong methods.

Numerical Solution of Algebraic Riccati Equations

Efficient algorithms for solving large-scale algebraic Riccati equations in optimal control and matrix equations applications.

Scalable Algorithms for Graph Problems

Development of parallel and distributed algorithms for spectral clustering, graph partitioning, and network analysis on massive graphs.

Numerical Methods for Maxwell''s Equations

High-order discretization and time-stepping schemes for electromagnetic wave problems including edge elements and structure-preserving integrators.

Implicit-Explicit Time Stepping for Multiscale Systems

Design of IMEX schemes that handle fast and slow components efficiently with asymptotic-preserving properties.

Numerical Methods for Chemotaxis and Nonlinear Diffusion

Computational techniques for solving chemotaxis equations and other nonlinear diffusion systems with positivity and unconditional stability.

Hierarchical Tucker Tensor Formats

Algorithms for constructing and manipulating hierarchical Tucker decompositions with applications to high-dimensional parameter spaces.

Finite Element Error Analysis for Nonlinear Problems

Development of rigorous a posteriori error estimates and error bounds for nonlinear FEM approximations including nonconforming elements.

Numerical Methods for Variational Inequalities

Computational algorithms for solving variational inequalities and complementarity problems with applications to obstacle and contact problems.

Parallel-in-Time Integration Schemes

Development of parallel time-stepping algorithms like parareal and multigrid reduction in time for massively parallel computing.

Data-Driven Discovery of Dynamical Systems

Sparse identification algorithms and symbolic regression methods for inferring governing equations from observational and simulation data.

Numerical Methods for Nonlinear Wave Equations

High-order schemes for nonlinear wave equations preserving energy, momentum, or symplectic structure over long integration times.

Iterative Methods for Nonlinear Equations

Development and analysis of Newton-Krylov methods, Jacobian-free techniques, and quasi-Newton methods for large-scale nonlinear systems.

Numerical Methods for Crystal Growth and Phase Fields

Computational techniques for phase-field models, dendritic growth, and pattern formation in crystal solidification.

Flux Reconstruction and Correction Procedures

High-order finite difference and finite element methods based on flux reconstruction and correction procedures for conservation laws.

Numerical Methods for Birth-Death-Mutation Processes

Algorithms for simulating branching processes and chemical reaction networks with applications to population dynamics and genetics.

Preconditioned Iterative Methods for Electromagnetics

Development of effective preconditioners for linear systems arising from edge element and nodal element discretizations of Maxwell equations.

Numerical Bifurcation Analysis in Partial Differential Equations

Computational methods for detecting and tracking bifurcations in infinite-dimensional systems arising from PDEs.

Hybrid Finite Element and Boundary Element Methods

Coupling strategies combining FEM for bounded domains and BEM for unbounded domains with applications to scattering and wave problems.

Numerical Methods for Long-Time Behavior Analysis

Algorithms for accurately capturing long-time dynamics, attractors, and asymptotic behavior without resonance or phase errors.

Gradient-Enhanced Surrogate and Emulator Models

Construction of surrogate models incorporating gradient information for improved accuracy in reduced-order modeling and sensitivity analysis.

Numerical Methods for Viscoelastic Materials

Computational techniques for modeling and solving problems in viscoelastic mechanics including rheological models and stress relaxation.

Mesh Adaptation and Metric-Based Refinement

Automated mesh adaptation algorithms using error estimates and metric tensors for optimal mesh configuration in adaptive computations.

Nonconforming and Mixed Finite Element Spaces

Development and analysis of nonconforming and mixed finite element discretizations for elliptic and parabolic problems with applications to incompressible flow and elasticity.

Exponential Integrators for Stiff Systems

Construction and analysis of exponential time integration schemes for highly stiff differential equations arising in reaction-diffusion and wave propagation problems.

Adaptive Mesh Refinement Strategies

Development of dynamic mesh adaptation algorithms using error estimators and solution features for efficient resolution of multiscale phenomena.

Numerical Methods for Fractional Calculus

Design and analysis of finite difference and spectral schemes for fractional differential equations with memory effects and anomalous diffusion.

Structure-Preserving Schemes for Gradient Flows

Development of energy-stable and entropy-stable discretizations for gradient flow systems in materials science and biological applications.

Operator Learning and Neural Operators

Research on learning infinite-dimensional operators from data using neural networks to accelerate the solution of parametric families of PDEs.

Hybrid High-Order Methods

Analysis and application of hybrid high-order discretizations combining high polynomial degrees with reduced coupling for complex geometries and multiphysics problems.

Stabilized Methods for Advection-Dominated Transport

Design of stabilization techniques including streamline upwind and residual-based methods for advection-dominated convection-diffusion problems.

Localized Orthogonal Decomposition Methods

Development of multiscale numerical methods using local basis functions for heterogeneous material problems and elliptic equations with oscillatory coefficients.

Conservative Schemes for Multiphase Flows

Development of mass-conservative and momentum-conservative discretizations for compressible and incompressible multiphase flow simulations.

Galerkin Orthogonal Polynomial Methods

Application of orthogonal polynomial bases and spectral Galerkin formulations for solving differential equations with high accuracy and efficient computation.

Numerical Homogenization via Asymptotic Expansion

Computational techniques for extracting effective material properties and macroscopic behavior from oscillatory microscale problems using asymptotic methods.

Fast Solvers for Electromagnetic Wave Problems

Development of efficient preconditioners and iterative solvers for large-scale discretizations of Maxwell equations in heterogeneous media.

Implicit-Explicit Time Stepping Methods

Analysis and implementation of splitting schemes that treat stiff and nonstiff components separately to achieve computational efficiency.

Tensor Networks for High-Dimensional PDEs

Exploitation of tensor decomposition structures for efficient storage and computation of solutions to high-dimensional parametric differential equations.

Numerical Methods for Nonlocal Diffusion

Development of finite element and finite difference schemes for integro-differential equations arising in peridynamics and anomalous transport.

Weak Galerkin and Discontinuous Weak Galerkin Methods

Novel discretization frameworks using weak derivatives and weak continuity to achieve stability and flexibility for diverse PDE systems.

Data-Driven Discovery of Governing Equations

Computational methods for identifying differential equations and constitutive laws directly from observational data using sparse regression and symbolic methods.

Robust Optimization and Worst-Case Design

Development of numerical algorithms for optimization problems under uncertainty with guarantees on solution quality across all feasible parameter realizations.

Unfitted Finite Element Methods

Design of discretization schemes operating on fixed meshes independent of domain boundaries using ghost penalty and enrichment techniques.

Numerical Analysis of Reaction-Diffusion Systems

Mathematical analysis and computation of spatiotemporal pattern formation, traveling waves, and bifurcations in coupled reaction-diffusion equations.

Block Krylov Subspace Methods

Development of Krylov subspace techniques for solving matrix equations and multiple right-hand side systems with improved computational efficiency.

Finite Element Methods for Optimal Control Problems

Analysis of discretization schemes for constrained optimization of systems governed by differential equations with stability and convergence guarantees.

Numerical Methods for Stochastic Partial Differential Equations

Design and analysis of temporal and spatial discretizations for SPDEs including strong and weak convergence analysis under various noise models.

Virtual Element Methods for Nonlinear Problems

Extension of virtual element methodology to nonlinear problems including contact mechanics, plasticity, and nonlinear diffusion with arbitrary polygon meshes.

Matrix-Free Algorithms for Scientific Computing

Development of computational methods that avoid explicit matrix assembly through operator evaluation for efficient large-scale parallel scientific simulations.

Numerical Methods for Coupled Multi-Field Problems

Discretization and solution strategies for strongly coupled systems involving mechanics, heat transfer, electromagnetism and chemistry.

Isoparametric Finite Element Analysis

Advanced analysis of curved element discretizations and geometric approximation errors in finite element methods for complex computational domains.

Numerical Methods for Population Dynamics Models

Computational techniques for age-structured, stage-structured and spatially heterogeneous population models including birth-death processes and evolution equations.

Fast Transforms and Sampling on Manifolds

Development of efficient algorithms for function approximation and sampling on manifolds with applications to geometric data analysis and manifold learning.

Numerical Methods for Free Boundary Problems

Computational approaches for Stefan problems, moving boundary problems and variational inequalities using implicit, explicit and adaptive techniques.

Singularity Extraction and Subtraction Methods

Techniques for handling singular behavior through analytical or numerical singularity separation to improve accuracy and convergence in computational methods.

Numerical Continuation and Bifurcation Analysis

Algorithms for tracking solution branches, locating bifurcation points and analyzing stability in parameterized nonlinear dynamical systems.

Mimetic Discretization Methods

Construction of numerical schemes that preserve key mathematical properties such as divergence-free solutions, discrete conservation laws and dispersion relations.

Numerical Methods for Kinetic Theory

Computational techniques for Boltzmann equations and kinetic transport including collision operators, asymptotic preserving schemes and moment closures.

Least Squares Methods and Petrov-Galerkin Formulations

Development of least squares finite element methods and nonsymmetric Petrov-Galerkin formulations for robustness in difficult PDE regimes.

Numerical Methods for Viscoelastic Flow

Discretization strategies for non-Newtonian fluids with memory effects including constitutive models and coupled momentum-stress systems.

Efficient Sampling for Bayesian Inference

Development of advanced MCMC methods including adaptive sampling, delayed acceptance and likelihood-free inference for inverse problems and calibration.

Coupled Electromagnetic and Thermal Simulations

Computational methods for multiphysics problems combining Maxwell equations with heat conduction and thermal effects in materials and devices.

Numerical Methods for Long-Time Integration

Development of schemes with superior long-term stability and accuracy for simulating multiscale temporal dynamics and climate-scale phenomena.

Space-Time Methods for Time-Dependent PDEs

Discretization techniques that treat space and time symmetrically within a unified framework enabling parallel computation and efficient adaptivity.

Gradient Descent and Optimization on Manifolds

Development of first-order and higher-order optimization algorithms that respect manifold structure for machine learning and inverse problems.

Numerical Methods for Polymer Dynamics

Computational approaches for macromolecular simulations including chain conformations, entanglement effects and non-equilibrium relaxation dynamics.

Multilevel Monte Carlo Methods for Uncertainty

Hierarchical sampling techniques that exploit multiple discretization levels to achieve improved computational efficiency in uncertainty quantification.

hp-Finite Element Methods

Theory and implementation of adaptive finite element methods combining mesh refinement with polynomial degree adjustment for exponential convergence.

Numerical Methods for Electrorheological Fluids

Computational schemes for coupled electromagnetic-mechanical problems involving field-dependent viscosity and smart fluid behavior.

Conservative Discretizations for Wave Equations

Development of schemes preserving energy and wave dispersion characteristics for accurate long-time simulation of acoustic and elastic wave propagation.

Numerical Methods for Surfactant Transport

Computational approaches for coupled surface-bulk transport with applications to interfacial tension, capillary flow and emulsion dynamics.

Parallelizable Multigrid Methods for Exascale Computing

Development of multilevel solvers designed for extreme-scale parallel architectures with communication-avoiding and asynchronous communication patterns.

Hybridizable Discontinuous Galerkin Methods

Development and analysis of HDG methods that reduce global degrees of freedom while maintaining high-order accuracy for elliptic and parabolic problems.

Exponential Integrators for Stiff Systems

Design of efficient exponential time integrators that handle stiff differential equations with improved stability and accuracy properties.

Graph Neural Networks for Scientific Computing

Application of graph-based deep learning architectures to learn and predict solutions of partial differential equations on unstructured domains.

Variational Data Assimilation and 4D-Var

Development of variational methods for combining observational data with numerical models in time-dependent inverse problems.

Finite Volume Methods for Nonconservative Systems

Numerical discretization techniques for hyperbolic systems with nonconservative products and their well-posedness analysis.

Kernel Methods and Radial Basis Functions

Theory and algorithms for meshfree interpolation and approximation using kernel functions with applications to PDE solving.

Geometric Integration and Lie Group Methods

Construction of time integration schemes that preserve geometric properties of differential equations on manifolds.

Adaptive Cross Approximation and Data Sparsity

Low-rank matrix approximation techniques for hierarchical computation and compression of large-scale dense matrices.

Entropy Stable Schemes for Hyperbolic Systems

Development of numerical methods that guarantee discrete entropy inequality for conservation laws and hyperbolic systems.

Mixed and Hybrid Finite Element Formulations

Construction of finite element spaces satisfying inf-sup conditions for coupled multi-field problems and saddle-point systems.

Surrogate Modeling and Metamodel Construction

Development of fast approximate models of expensive computational simulations for design optimization and sensitivity analysis.

Numerical Methods for Fractional PDEs

Discretization schemes and solvers for partial differential equations involving fractional derivatives and integro-differential operators.

Multifidelity and Multiscale Surrogate Methods

Techniques combining models of varying computational cost and resolution to accelerate optimization and uncertainty quantification.

Time Integration for Allen-Cahn Equations

Energy-stable and highly accurate temporal schemes for phase-field models with application to complex interface dynamics.

Data-Driven Reduced Basis Methods

Construction of low-dimensional subspaces from snapshots or experimental data for rapid approximation in parametric settings.

Numerical Methods for Kinetic Equations

Discretization and solution algorithms for Boltzmann and related kinetic equations with asymptotic-preserving properties.

Structure Preservation in Optimization Algorithms

Design of optimization methods that respect mathematical structure of feasible sets and constraint manifolds.

Neural Operator Learning and DeepONet

Machine learning approaches to learn operators mapping between function spaces for parametric PDE families.

Discontinuity Resolving Schemes and Shock Capturing

Numerical methods that accurately resolve sharp gradients and shocks without artificial oscillations in hyperbolic problems.

Convex Optimization and First-Order Methods

Analysis and development of gradient-based algorithms including acceleration, proximal methods, and splitting strategies.

Numerical Continuation and Bifurcation Analysis

Computational techniques for tracking solution branches of nonlinear equations and detecting bifurcation points.

Parametric Uncertainty in Computational Models

Methods for quantifying propagation of input parameter uncertainties through complex computational models and simulations.

Operator Splitting and Alternating Direction Methods

Decomposition techniques for splitting complex operators into simpler subproblems with improved computational efficiency.

Nonlinear Wave Equations and Solitary Waves

Numerical methods for nonlinear hyperbolic equations and construction of localized traveling wave solutions.

Surrogate-Assisted Evolutionary Algorithms

Integration of fast surrogate models with evolutionary optimization for expensive multi-objective design problems.

Stabilized Methods for Advection-Dominated Transport

Numerical schemes using stabilization techniques to suppress spurious oscillations in convection-dominated problems.

Coarse Space Design for Domain Decomposition

Development of scalable preconditioners through optimal construction of coarse problem spaces in parallel computing.

Orthogonal Polynomials and Spectral Approximation

Theory and application of orthogonal polynomial bases for high-order approximation with exponential convergence.

Augmented Lagrangian and Penalty Methods

Constrained optimization algorithms using augmented objectives and penalty formulations for complex constraint handling.

Quasi-Monte Carlo Methods and Low Discrepancy Sequences

Deterministic sampling techniques for numerical integration in high dimensions with superior convergence rates.

Finite Element Assembly and Matrix-Free Methods

Efficient computational techniques for avoiding explicit matrix construction in finite element implementations.

Machine Learning for Surrogate Metamodels

Use of supervised and unsupervised learning algorithms to construct fast approximations of expensive simulations.

Coupled Multiphysics and Multidomain Solvers

Numerical methods for solving coupled systems of equations from different physical domains with appropriate interface conditions.

Higher-Order Time Discretization Schemes

Construction of temporal integration methods achieving high-order accuracy with optimal stability regions and error constants.

Finite Element Methods with Local Enrichment

Extended and generalized finite element methods using enrichment functions for capturing localized features without mesh refinement.

Uncertainty Propagation via Polynomial Chaos

Spectral stochastic methods using orthogonal polynomial expansions for efficient uncertainty quantification in parametric systems.

Fully Discrete Analysis and Convergence Theory

Rigorous mathematical analysis of combined spatial and temporal discretization errors for complex numerical schemes.

Greedy Algorithms in Model Order Reduction

Efficient basis selection and snapshot sampling strategies for constructing optimal reduced-order approximations.

Monotone Schemes and Maximum Principles

Development of numerical methods that preserve monotonicity and discrete maximum principles for diffusive and transport problems.

Staggered and Collocated Grid Arrangements

Study of different variable placement strategies in finite difference and finite volume methods for coupled systems.

Geometric Deep Learning for Manifold-Valued Data

Development of neural network architectures that respect intrinsic geometric structures and differential constraints on nonlinear manifolds for scientific computing applications.

Certified Numerics and Rigorous Error Bounds

Construction of computer-assisted proofs and validated numerical methods that provide mathematically rigorous bounds on computational errors in nonlinear problems.

Asymptotic-Preserving Numerical Schemes

Construction of schemes that accurately capture asymptotic limits of singular perturbation problems without resolving small scales.

Adaptive Cross Approximation and Low-Rank Compression

Advanced matrix and tensor compression techniques using hierarchical and cross-based methods for efficient computation with extremely high-dimensional and dense data structures.

Error Bounds and Guaranteed Numerics

Rigorous computation of verified error bounds and certified solutions for deterministic and stochastic problems.

Coupled Multirate Time Integration

Temporal schemes using different time steps for different system components to exploit disparity in dynamics.

Time Integration for Multi-Physics Coupling

Development of high-order splitting and decoupling schemes that efficiently handle multiple physical processes operating at different temporal scales without sacrificing stability.

Numerical Linear Algebra for Eigenvalue Problems

Development of efficient algorithms for computing eigenvalues and eigenvectors of large sparse matrices.

Computational Harmonic Analysis and Signal Processing

Construction of efficient transforms, frame theory, and sampling strategies for decomposing and analyzing complex signals arising in scientific and engineering applications.

Graph Neural Networks for Differential Operators

Learning universal approximators for nonlinear differential operators and functional mappings using message-passing architectures on unstructured computational graphs.

Parareal and Parallel-in-Time Methods

Development of time-parallel algorithms that decompose temporal evolution to achieve massive parallelization for solving time-dependent multiscale and multiphysics problems.