Projects
Sentiment Analysis on Amazon Customer Reviews
A comprehensive Bachelor's Thesis project exploring how machine learning enhances customer experience through sentiment analysis. Analyzes Amazon reviews (2015-2020) using multiple ML algorithms including Logistic Regression, Decision Trees, and VADER lexicon-based approaches, achieving 93% accuracy in sentiment classification.
Daily Bulletin Scrapers
A Python-based scraping system that automatically downloads and parses daily financial bulletins from major Turkish brokerage firms. Uses PDF processing, Selenium automation, and fuzzy matching to extract company news and BIST ticker symbols, aggregating data into structured CSV format for financial analysis.
Endoscopy Triage & Diagnosis Tool
A specialized Shiny web application for clinical decision support in gastrointestinal bleeding cases. Uses Random Forest machine learning models to predict bleeding source (upper/mid/lower GI), urgent endoscopy needs, and patient disposition (ICU vs. non-ICU), providing probabilistic outputs with feature importance visualization for clinicians.
Real Estate Webform Auto Filler
A specialized Selenium-based automation script for real estate property assessment forms. Automates complex multi-section webforms including building details, classifications, heating systems, and property dimensions. Uses dynamic waiting, error handling, and ChromeDriver management to efficiently process property data with high reliability.