👋🏻 My name is Jordan

📝 Master of Information and Data Science candidate at UC Berkeley

🌏 A human-centered problem solver, approaching data science as a technical challenge with a focus on people-oriented outcomes

🔭 I’m currently learning about transformer architectures for NLP, mitigating model bias, and time series analysis

Language endangerment threatens cultural diversity, with over 40% of the world’s 7,000 languages at risk of disappearing. As these languages vanish, so do the unique histories, identities, and perspectives they embody, often from marginalized groups. This project aims to predict a language’s endangerment level, from extinct to not endangered, using features such as speaker counts, political status, and urbanization and internet usage rates in the countries where it is spoken. By applying multiple predictive models, we seek to identify the key factors driving language decline and enable earlier, targeted preservation efforts.

Diabetic retinopathy (DR) is a diabetes-related eye disease that damages the blood vessels in the retina, potentially leading to vision loss or blindness. This project aims to develop a convolutional neural network (CNN) for detecting diabetic retinopathy (DR) from retina images captured through fundus photography under diverse imaging conditions. To enhance the model's generalization and minimize overfitting, the approach incorporates image transformation and data augmentation techniques, enabling the system to perform robustly across a wide range of visual inputs.

In digital commerce, personalization and product bundling are widely used strategies aimed at increasing consumer engagement and spending. This study investigates whether identity-based personalization enhances the effectiveness of product bundling in driving purchase behavior. By using a 3x2 factorial design, users are assigned to one of six simulated purchase environments. By offering a maximum $300 prize, we attempt to motivate participants to treat the simulation as close to reality as possible. Our hypothesis states that bundling and personalization will individually increase intent to purchase amounts, while the combination of both will also result in higher purchase amounts.

Utilizing open source data from the New York Metropolitan Museum of Art, our team converted the museum's object collection data into a graph database structure and used graph algorithms to uncover key insights that inform exhibition ideas, object curation, and user experience enhancements. We modeled the data in Neo4j and used Pagerank, Louvain Modularity, and Closeness Centrality algorithms. Finally, we developed exploratory plots and recommendations based on centrality scores and community detection.

An exploratory data analysis project that uncovers relationships between gender, employment, and primary school enrollment at a global scale. By using Development Indicators and economic categorizations from the World Bank, my team discovered correlations between a country's female primary school enrollment rates and employment rates. By filtering and reshaping the data using melting in Python, we were able to analyze indicators for our chosen variables accross the years from 2017-2023, handling missing data with consistency and producing high-quality visualizations.