Skills & Projects
Technical Skills
| Category | Details |
|---|---|
| Programming | Python (PyTorch, TensorFlow, scikit-learn, Hugging Face Transformers, pandas, NumPy, matplotlib, seaborn), C++, Perl, SQL |
| Tools | Git, HTML, LaTeX |
| Mathematical Software | Macaulay2, Mathematica, MATLAB |
Projects
SAND: Sheaf Analysis of Network Data
Georgia Tech Research Institute
Applied sheaf-theoretic methods to model and analyze network traffic structure, building a mathematical framework for identifying anomalous patterns indicative of cyber threats. Analysis involved creating feature vectors for rolling-time windows over the course of one day’s activity and then passing these vectors through an autoencoder. Experimental results showed significantly higher autoencoder loss during red team attacks compared with benign data.
Cross-Modal Image-Caption Retrieval
University of Nebraska-Lincoln
Proposed, designed, and implemented a Correlational Autoencoder for cross-modal image-caption retrieval using pretrained ResNet-50 and BERT feature extractors with a shared 512-dimensional latent space. Trained and evaluated three architectural variants on the Flickr8k dataset; the final contrastive alignment model achieved Recall@10 of 50.4%, a 2x improvement over the MSE baseline.
Education Highlights
- PhD in Mathematics, University of Nebraska-Lincoln (expected December 2026)
- Relevant coursework: CSCE 879: Introduction to Deep Learning
- Advised by Eloísa Grifo
- MS in Mathematics, Missouri State University, 2021
- BS in Applied Mathematics, Missouri State University, 2019
- Minor in Computer Science (courses in Database Systems and Machine Learning)
- Minor in Spanish