Personal Work and Projects

Does Gradient Descend Have Better Ball Knowledge than Me? Python · Scipy, Numpy

Comparison of parameter estimation via linear regression and neural networks for football match result predictions. Achivied within-noise RSP of 0.175 (close to prediction markets) and showed thar a linear model is more interpretable and robust compared to an MLP.

View on GitHub → Read here →

Physics-Informed DeepL for Binary Black Hole Mergers Python · PyTorch

An empirical study of whether binary-black-hole inspiral waveforms admit a Hamiltonian latent-space representation, using a physics-informed autoencoder.

View on GitHub →

Empirical Study of Numerical Solutions to the Black-Scholes-Merton PDE Python · Scipy, Numpy

An empirical study of 4 numerical methods on the BSM PDE, focusing on compute time, accuracy, and derivative estimates.

View on GitHub →