Probabilistic Neural Operator Financial Experiment: SPY Data, Python Implementation, Numerical Results, and Figures

Abstract

This repository contains the computational materials supporting the numerical experiments presented in the associated research manuscript on probabilistic neural operators. The repository provides the Python implementation used to construct and evaluate the finite-level probabilistic neural operator, perform the representation-level and Monte Carlo sampling convergence experiments, and generate the numerical results and publication-quality figures reported in the study. The empirical financial experiment uses historical SPY (SPDR S&P 500 ETF Trust) observations obtained from Yahoo Finance. The downloaded adjusted-price data are processed into daily log returns and subsequently used to construct the functional input-output pairs for the probabilistic operator experiment. The computational workflow was executed in Google Colab using Python and the scientific-computing libraries specified in the accompanying source code. The repository includes the Python source code, the SPY data used in the experiment, numerical result tables in CSV format, and publication-quality vector PDF figures. The code and accompanying materials are provided to facilitate reproducibility of the reported numerical experiments. The SPY financial observations were obtained from Yahoo Finance and remain subject to the terms and conditions applicable to that source. The repository does not claim ownership of the underlying financial data.