Quantitative Researcher | Algorithm Engineer | Applied Mathematician

I bridge the gap between complex stochastic theories and high-performance numerical implementations. I am passionate about extreme performance optimization, mathematical modeling, and pushing computational limits.

• Numerical Optimization & DEQs: Developed a high-precision solver for Cascaded Monotone Equilibrium Networks (MonDEQs) from scratch using PyTorch (Double Precision).

  • Achieved $10^{-14}$ absolute precision, proving numerical closure for coupled monotone inclusion problems.
  • Realized a $1588\times$ computational speedup by discovering an Inexact Splitting regime ($k=1$) that maintains global stability without local convergence.
  • Identified and suppressed the "Dead Zone" latency (750+ iterations) inherent in sequential solvers, enabling Anytime Convergence for real-time neuromorphic systems.
  • The core theoretical framework and convergence proofs are currently under review for a top-tier international conference in mathematical systems theory (MTNS 2026).

• Continuous-Time Dynamics Modeling (Denison Research): Designed deterministic ODE frameworks to model population flux and state transitions (applied to Liver Fibrosis).

  • Solved high-dimensional linear ODE systems via Matrix Exponentials, eliminating numerical stepping errors and guaranteeing exact probability mass conservation for irregularly sampled longitudinal data.
  • Engineered a custom Maximum Likelihood Estimation (MLE) optimization pipeline; applied AIC/BIC criteria to rigorously penalize model complexity and prevent overfitting on sparse transition matrices.
  • Handled observational noise and diagnostic ambiguity by embedding Mixture Likelihoods (treating uncertain observations as latent probabilistic states).
  • Extended the analytical framework to non-autonomous systems $\mathbf{Q}(t)$ using RK45 numerical integration to simulate counterfactuals under dynamic, time-varying covariates.

• Quantitative Portfolio Construction & Risk Management: Architected an institutional-grade, Object-Oriented (OOP) toolkit for Strategic Asset Allocation (SAA) and model stress-testing.

  • Developed a robust Mean-Variance Optimisation engine utilising Sequential Least Squares Programming (SLSQP) to handle complex, real-world institutional mandates (e.g., asymmetric bounds, zero-weighting, and sector-level concentration caps).
  • Implemented an advanced Model Fragility Testing framework to evaluate optimizer behavior under stressed and economically unsound correlation regimes.
  • Engineered a numerical repair module using Eigenvalue Decomposition and Clipping to project non-Positive Semi-Definite (non-PSD) covariance matrices back onto the PSD cone, ensuring mathematical validity for corrupted market inputs.
  • Conducted dual-metric evaluations contrasting pure mathematical efficiency (Sharpe Ratio) against investor risk preferences (Exponential Utility functions under CARA assumptions).

• Stochastic Dynamics & Complex Systems Modeling: Modeled the impact of intrinsic noise on the reliability of bursting rhythms in high-dimensional biophysical networks (CA1 Pyramidal Cells).

  • Architected a 5D non-linear Stochastic Differential Equation (SDE) framework with multiplicative noise (square-root diffusion) to capture state-dependent volatility.
  • Engineered a decoupled Monte Carlo simulation engine, featuring a custom Euler-Maruyama solver and systematic deterministic ablation studies (factor knockouts), to identify dynamic bifurcations.
  • Quantified temporal reliability degradation by extracting event-driven signals (e.g., inter-burst intervals) across high-volume simulated paths and evaluated system resilience using statistical metrics (Coefficient of Variation).

• Mathematics: Stochastic Calculus, Measure Theory, Nonlinear Dynamics, Convex Optimization.
• Programming: Python (PyTorch, NumPy, SciPy, Pandas), R, MATLAB, MySQL.
• Tools: LaTeX, Git, Linux, XPPAUT.

• Mandarin & Gan: Native proficiency.
• English: Full Professional Proficiency (C1/C2).
• German: Currently learning (A2).

• Bachelor of Science in Mathematics, Financial Mathematics & Statistics.
• Focus: Stochastic Processes, PDE and Waves, Mathematical Computing, Measure Theory and Fourier Analysis(Lp & Hilbert Space), Analysis (Real & Complex), Data Structures and Algorithms, Financial Derivatives.
• Honors: Recipient of competitive research scholarships (Denison Research Scholar, Vacation Research internship) and academic achievement awards.

When I'm not pushing the limits of numerical solvers or analyzing stochastic processes, you can find me:

• Applied Signal Processing: Tweaking electric guitar effects pedals to dial in the perfect tone. I'm a massive fan of Shoegaze, Post Rock, Space and Psychedelic Rock, and an avid collector of rare first-press vinyls and CDs.
• Solo Off-the-Grid Backpacking: Exploring the world's most remote landscapes. I've trekked the Australian Red Centre, summited snow peaks in New Zealand, navigated the wintery wilderness of East Hokkaido, and reached the northernmost point of mainland UK in the Scottish Highlands.
• Strategy & Philosophy: Playing Go (Weiqi), composing traditional Chinese poetry, and exploring Daoist philosophy—finding the perfect equilibrium in both complex systems and life.

"For stochastic tomorrow."

Reach me at: [wuyefan718@gmail.com]