A group data mining project (DATA 3421) that asks: what separates a blockbuster from the rest? We defined "blockbuster" two ways — top 25% by revenue and top 25% by popularity — then trained parallel Decision Tree, Random Forest, Logistic Regression, and KNN classifiers on each. Feature engineering included log-transformed budgets, Bayesian-adjusted ratings, ROI, genre complexity signals, and interaction terms. A final comparison table contrasts model accuracy across both targets.

This project involved building a classification model to predict survival outcomes for Titanic passengers. I applied feature engineering techniques (handling missing values, encoding categorical variables), and used ensemble methods including Random Forest and Gradient Boosting. The model was evaluated using cross-validation and achieved strong accuracy on the Kaggle test set.

A front-end website that fetches real-time stock market news using a financial news API. Users can search for any stock ticker and get the latest headlines. Built entirely with HTML, CSS, and vanilla JavaScript, with a clean responsive layout and live data rendering.

Used Selenium WebDriver to automate browser-based tasks such as form filling, login flows, and data extraction. The project scrapes rental listings from Zillow and automatically fills out Google Forms with the house data (price, address, availability), producing a clean spreadsheet for quick decision-making.

Built a Python-based scraper using BeautifulSoup to extract the Billboard Hot 100 chart for a user-specified date. The top 100 songs are then automatically added to a new Spotify playlist using the Spotify API, letting you travel back in time through music.

A desktop quiz app built with Python and Tkinter. Users can answer True/False questions fetched from the Open Trivia Database API, track their score, and see results at the end. The app uses a clean GUI layout with two interactive buttons for each question.

A Python app that connects to a weather API to retrieve current conditions and forecasts for a given location. It then automatically sends a formatted daily weather summary via email using SMTP, presenting the data in a clean table format.

A Python and Tkinter app that shows Hindi flashcards for vocabulary learning. Cards flip automatically after 3 seconds to reveal the English translation. Progress is saved so already-learned words are skipped in future sessions, making it an efficient spaced-repetition tool.

A faithful recreation of the classic Pong arcade game built with Python's Turtle module. Two paddles, a bouncing ball, and a scoreboard — all implemented using object-oriented principles. Each game element (paddle, ball, scoreboard) is its own class. First player to 10 points wins.

The classic Snake game rebuilt from scratch in Python using the Turtle graphics library. The snake, food, and scoreboard each live in their own class, with clean separation of concerns. The game increases speed as the snake grows, and handles wall collisions and self-collisions for a complete gameplay loop.