projects

research

• Daytime Segmentation Approaches for Depression Classification - feature extraction for time segmentation in acceleroemeter data. publication for (hopefully) IEEE Journal of Biomedical and Health Informatics in progress

academic work

• Climate Change Forecasting with ARIMA - time series analysis of global temperature anomalies (1880-2024) using NASA GISTEMP data, projecting 1.13°C sustained anomalies through 2030
• Collection Score: Debt Recovery Prediction System - machine learning system predicting payment probability with 91.3% accuracy using LightGBM, integrated with Mistral AI for personalized collection strategies. projected 40% productivity increase for collection teams
• Cardiac Arrest Predictor - unsupervised learning solution identifying heart attack risk patterns using K-means and Mean-Shift clustering
• Infraestructure Impact Analysis - PCA-based infrastructure index for Mexican states, analyzing relationships between development, GDP per capita, and poverty indicators