Assessing Suicide Risk in Social Media Posts

HTML

NLP

Classified suicide risk in social media posts versus other mental health conditions (e.g., depression, anxiety) using transformer-based models (MentalBERT, DisorBERT) and the Reddit Mental Health Dataset, integrating pre- and post-COVID data for improved generalizability.

Transit in Beijing

HTML

Tableau visualizations

Analyze transportation behavior within Beijing's 5th Ring Road from 2007 to 2012, using the Geolife dataset from Microsoft Research Asia. Our study explores temporal, spatial, and compositional trends in transportation modes, providing insights for city planners on optimizing infrastructure and addressing travel patterns affected by seasons, holidays, and daily commuting habits.

Predicting Student Drop Out Rates

Machine Learning

Analyze the leading factors contributing to student dropout in higher education by developing and comparing four predictive models: baseline, logistic regression, random forest, and neural network.

Descriptive Regression Analysis

Agnostic Statistics

Answer the following: "What is the relationship between age and BMI, while taking into account other demographic and lifestyle factors?"

GAME OF THRONES NAMED ENTITY RECOGNITION NETWORK

Python

pyvis visualizations

Extract relationships between Game of Thrones Characters found from webScraping the offical wikipedia site. Through graph analyses (centrality measures/community detection) and visualization, via Pyvis, analyze the relationships between characters

CHATGPT API CHATBOT

Python

20 questions with Chatbot, informed by more than 230,000 Reddit posts and comments

An ask-me-anything chatbot using ChatGPT API that answers all your questions about data science, ML, and AI based on subreddit content.