Put double rings on some integer lattice points in the first quadrant of the cartesian plane. From bottom to top, left to right, if coordinate (x,y) and (x,y+1) are have any circles, then put a simple circle on (x+1,y+1). What you obtain is a chinampa. To count certain polycrhony groups we introduce the theory of chinampas and study their enumeration problem.
"CheMLFlow: An Open-Source Platform for Cheminformatics and Materials Informatics Applications" link, with Brendan Smith, Susana Lopez-Moreno, Sangil Kim, Jose L. Mendoza-Cortes, and Nijamudheen Abdulrahiman. We propose a ML pipeline compatible with agentic coding. This project is a step towards reducing the reproducibility problem in science and reducing problems with the use of AI agents. While focused on Cheminformatics, it can be applied to other areas of AI.
"PCA-Enhanced Adaptive NVAR Framework for High-Resolution Sea Surface Temperature Forecasting in the East Sea" link, with Azimov Sherkhon, Susana Lopez-Moreno, JinYong Choi, and Sangil Kim. We modify our Adaptive-NVAR to estimate sea surface temperature.
"Adaptive Nonlinear Vector Autoregression, Robust Forecasting for Noisy Chaotic Time Series" link, with Azimov Sherkhon, Susana Lopez-Moreno, Sieun Lee, and Sangil Kim. We create a neural network (based on NVAR) to estimate time series from caotic dynamical systems with noise.
"Standardization of Post-Publication Code Verification by Journals is Possible with the Support of the Community" link, with Susana Lopez-Moreno and Sangil Kim. To fight the reproducibility problem in machine learning we explore a pipeline for post-publication code verification that involves journals, reviewers, students/researchers and job interviewers.