Selected Projects in Machine Learning, Cloud Computing, Relational Databases, and Statistics

Big Data Analysis of Renewable Energy in Emerging Economies

Leveraging the power of Google Cloud Platform, this big data project examines the impact of socio-economic factors on renewable energy consumption and solar radiation potential in emerging economies. Advanced data analytics tools, including Google Bigtable and Python, are employed to uncover how socio-economic trends shape renewable energy use, offering key insights for sustainable energy policies in these rapidly evolving markets.

View code on Github

Optimizing Disease Diagnosis on Amazon Web Services

This research tackles feature redundancy in large medical datasets by using ensemble learning techniques on a cloud platform. Big data techniques, genetic algorithms for feature selection, and deep learning enhancements are employed. The optimized model is hosted on AWS for scalability and accessibility.

View code on Github

Relational Database Road Safety Web Application for Texas Drivers

Drive Smart Texas is a web app promoting road safety in Texas by providing users with tailored historical accident data. We processed a vast traffic incident dataset, used MySQL Workbench for storage, and deployed seamlessly on Heroku via ClearDB. The platform offers route hazard insights and allows users to report new incidents.

View the app Drive Smart Texas
View code on Github

Linking a Network of Indiana Technologies

In collaboration with AnalytiXIN, this project analyzes clinical trial data in Indiana, comparing it with national trends to address equitable access to clinical trials. Utilizing data from Clinicaltrials.gov, U.S. Census, and CMS, the study highlights discrepancies in trial distribution and identifies key socio-economic factors influencing trial locations. This insightful analysis is pivotal for strategizing improvements in clinical trial accessibility across diverse regions of Indiana.

View code on Github

Advanced Object Detection: CaDOd Project

This innovative project in computer vision advances object detection, focusing on accurately identifying cats and dogs in images. It leverages a multi-phase approach, incorporating techniques like EfficientDet models, YoloV8, and PyTorch-based neural networks, to refine object detection accuracy significantly.

View code on Github

Utilizing advanced statistical methods in SPSS, this gun ownership policy memo provides a comprehensive data-driven analysis of the correlation between gun ownership and various demographic, educational, and political variables in the U.S. Methods used include descriptive statistics, independent samples t-tests, correlation analysis, and multivariate logistic regression.

View the policy memo

Skills-based Projects

A selection of smaller projects demonstrating specific data science and ML skills.