Projects

Selected work in scientific machine learning, manifold learning, numerical PDEs, data assimilation, and optimization. The featured projects below are the best starting points.

Featured Projects

Neural Operators for 1D Committor Learning

Machine Learning Operator Learning

A comparison of DeepONet and FNO for learning the committor operator across parametric and Gaussian-random-field potential families.

View Project

Diffusion Maps for Committors and Molecular Manifolds

Machine Learning Manifold Learning

A data-driven geometry project using TMDmap for committor estimation and diffusion maps for symmetry-aware molecular embeddings in LJ7.

View Project

DiffusionNet for Out-of-Sample Molecular Embeddings

Machine Learning Manifold Learning

A learned encoder for extending diffusion-map coordinates to unseen LJ7 configurations and dense MALA trajectory data.

View Project

Other Projects

MLB Fantasy Sports Project: Preseason 2025 Hitter Value Forecasting

Machine Learning Data Science

A data science project for MLB fantasy sports, using machine learning to forecast hitter value in the 2025 preseason.

View Project

Concurrent Game of Life Cycle

Parallel Computing Numerical Methods

This project uses Conway's Game of Life as a concurrency case study. Each cell update depends only on its local neighborhood.

View Project

Variational Neural Network for the Committor

Machine Learning Numerical PDEs

A variational neural solver for the rugged Muller committor with Monte Carlo validation near the \(q=0.5\) transition surface.

View Project

PINNs for the Allen-Cahn Equation

PINNs Numerical PDEs

A physics-informed neural network for the 1D Allen-Cahn equation, benchmarked against a spectral reference solution.

View Project

Spectral Baseline for Allen-Cahn

Spectral Methods Numerical PDEs

A Fourier spectral reference solver used to evaluate the Allen-Cahn PINNs model across the full space-time grid.

View Project

Optimal Interpolation for Point Vortex System

Data Assimilation Numerical Methods

An optimal interpolation study for chaotic point-vortex dynamics using drifter observations and covariance diagnostics.

View Project

Boundary Value Problem (BVP) Solver

Finite Elements Numerical PDEs

A custom finite-element workflow for a 2D elliptic boundary value problem with mesh generation and stiffness construction.

View Project

Spring System Optimization

Optimization Numerical Methods

A constrained spring-network energy minimization project comparing gradient descent and conjugate-gradient behavior.

View Project

More to come...

More projects will show up here over time.

* Still no promises on this... unless you count promises to procrastinate!