This portfolio showcases my work in data analysis and UX design, featuring projects that leverage machine learning, data visualization, and user-centered design to transform complex data into intuitive, socially impactful solutions.
This project examines global wealth concentration (Top 10% and Bottom 50% shares) between 2020 and 2024. Using data from the World Inequality Database and the World Bank, it identifies two distinct mechanisms of inequality: elite concentration in emerging markets and high household indebtedness in financialized advanced economies. The study reveals that wealth exclusion is a structurally embedded global phenomenon that does not automatically diminish with national income growth. more
This project analyzes U.S. cancer incidence, mortality, and survival data to build an interactive Tableau dashboard inspired by Cancer Facts & Figures 2025. By examining geographic patterns, mortality-to-incidence ratios (MIR), and healthcare system performance, it highlights disparities and progress across states, cancer types, and sexes. The dashboard translates complex public health statistics into accessible, exploratory insights for researchers, policymakers, and the public. more
This project analyzed 550 Amazon bestselling books (2009–2019) using SQL to identify trends in ratings, pricing, genre shifts, and author performance. It highlights what drives commercial success in the bestseller market, including the structural shift toward Non-Fiction and volume-driven growth. more
Image by Peter Mooney via Flickr (CC BY-SA 2.0)
This project investigates whether high-altitude environments are a causal driver of elite endurance running performance using Olympic marathon data. It utilizes both between-country and within-country analyses to distinguish physiological effects of altitude from cultural and socioeconomic factors. more
This project seeks to predict both heart disease and diabetes, filling a gap in existing models focused on individual diseases. Utilizing the 2022 CDC survey data from 400,000+ adults, it uncovers insights on physical status, regional health disparities, and correlations with chronic diseases. This project includes the development of a user-friendly website for predictions. more
The main aim of this project is to identify distinct vocalizations made by my cat, particularly focusing on meow sounds. Utilizing audio feature extraction and PCA, three clustering methods were employed for unsupervised learning. The final voting clustering process identified 11 distinct sounds unanimously agreed upon by all methods. more
The project aims to build an image classifier to distinguish between cat images from various breeds and personal photos of my cat, Mellow. Using transfer learning to boost performance due to limited data, the model achieved 100% training accuracy and 95% validation accuracy. more
This project conducts Exploratory Data Analysis (EDA) on animal shelter data to create a binary classification model predicting cat adoption. By identifying factors influencing adoption, it aims to improve care and support for animals. The final voting classifier model achieved a 77% accuracy and 81% F1 score for adoption prediction. more