Angela Diaz

Designed and implemented a semantic search system using sentence embeddings and FAISS vector indexing to efficiently retrieve relevant information from large collections of unstructured documents. document analysis. 

Built a deep learning model (CNN) for skin cancer classification using medical imaging data, incorporating preprocessing, augmentation, and Grad-CAM for model interpretability. 

Developed and evaluated time series forecasting models (ETS, ARIMA) to analyze trends and predict retail demand, supporting data-driven decision-making. 

Applied supervised machine learning techniques, including logistic regression and SVM, to classify obesity risk, evaluating model performance across multiple classes.