PROJECT ARCHIVE

Things I've worked on.

Data analytics, machine learning, deep learning, cybersecurity and a healthy amount of experimentation.

01

DATA ANALYTICS · CAPSTONE

Waco Downtown
Farmers Market

2025

MARKET ANALYTICS 2019—2024
DATA RANGE 5 YEARS
FOCUS SALES
PROGRAM SNAP

The question

What could five years of market data tell us about vendor participation, sales behavior, food-assistance programs and the overall operation of a community farmers market?

What I worked on

As team lead, I worked across a multidisciplinary group of data science and computer science students to analyze vendor sales, SNAP, Double Up Food Bucks, tokens, gift cards and participation patterns.

Python Data Cleaning Analytics Visualization Leadership
02

MACHINE LEARNING · CYBERSECURITY

Network Intrusion
Detection

ML

F1 SCORE 0.896

192.168.0.14 NORMAL

10.0.12.83 ATTACK

172.16.23.4 NORMAL

10.0.8.91 ATTACK

192.168.1.72 NORMAL

The goal

Develop a classification system capable of identifying malicious network traffic while preserving strong performance across classes.

The approach

I used RobustScaler during preprocessing and trained an XGBoost classifier, achieving an F1 score of approximately 0.896.

XGBoost RobustScaler Python Classification Cybersecurity
03

DEEP LEARNING · FINANCE

Financial Time-Series
Modeling

AI

GOOG MODEL COMPARISON

The experiment

Compared a range of traditional machine-learning and deep-learning architectures for financial prediction and time-series modeling.

Models explored

DNN, CNN, LNN, GRU, Transformer, LSTM, Extra Trees and XGBoost.

LSTM GRU Transformer CNN XGBoost
04

UNSUPERVISED LEARNING

Vehicle Clustering &
Dimensionality Reduction

ML

Cluster 01
Cluster 02
Cluster 03

The goal

Explore naturally occurring groups within vehicle data and understand how different dimensionality-reduction techniques reveal underlying patterns.

Methods

PCA, K-Means clustering, hierarchical clustering, t-SNE and UMAP.

PCA K-Means t-SNE UMAP Clustering
05

COMPUTER VISION · DEEP LEARNING

CNN Image
Classification

DL

7 🐸 3 🚗 9 🐱 5 ✈ 2 🐶 8 🚢

The work

Built and evaluated convolutional neural networks for image-classification problems across multiple benchmark datasets.

Datasets

MNIST, CIFAR-10 and CIFAR-100 with model training performed across extended epochs while monitoring performance metrics.

CNN Deep Learning MNIST CIFAR-10 CIFAR-100