Code

On my github you can explore some of my public projects I have been working on:

The IRNet optimization method: Repository can be found here. This project presents an optimization method designed to determine the optimal locations for deploying monitoring stations to reconstruct a physical quantity. Unlike classical machine learning models, this method incorporates statistical parameters, including the structure of the variance-covariance matrix.

The rankMax algorithm for multi-class monitoring networks: Repository can be found here. The rankMax algorithm is designed for reconstructing signals across multiple locations when information is limited to a small number of devices with varying capabilities. This project focuses on implementing a technology to enhance monitoring networks in such constrained scenarios.

The Robust Black Carbon Proxy Model: Repository can be found here. In this project, I analyze the critical steps involved in building a robust machine learning model for Black carbon prediction. The model addresses challenges related to data quality, such as imputation, denoising, and aggregation, to ensure reliable predictions even when high-end Black carbon measurement devices are unavailable.