This website contains a collection of my projects, each showcasing my passion for data analysis. These projects, recognized on Kaggle (bronze medals, over 15,000 views, 300+ forks), demonstrate my skills across various data analysis tools, programming languages, and problem-solving abilities. Each project is a unique journey, providing insightful solutions to real-world challenges.
- Conducted comprehensive analysis of more than 250k medical records, utilizing exploratory data analysis (EDA) and hypothesis testing to identify key factors impacting diabetes.
- Built and evaluated multiple predictive models for diabetes classification, using pipelines and resampling techniques to address class imbalance. The best-performing ensemble models achieved up to 86% accuracy and AUC values around 0.82, supporting early detection and improved management of diabetes.
- Analyzed 20k hotel reviews with advanced data analysis techniques, including sentiment analysis, prediction modeling, topic modeling, and negative reviews clustering.
- Achieved a high accuracy of 85% in sentiment prediction, providing deep insights into customer feedback to boost satisfaction. Identified key topics in reviews, allowing targeted improvements in services and facilities to meet customer expectations.
- Conducted comprehensive analysis of retail sales data in Istanbul (100k+ records), involving data cleaning, preparation, exploratory data analysis (EDA) to uncover meaningful patterns and insights, including the correlation between total revenue and currency rates.
- Developed highly accurate predictive models with an accuracy rate of 98%, resulting in enhanced sales predictions, optimized marketing strategies, and improved financial performance.
- Conducted in-depth analysis of US real estate data (200k+ entries) using SQL, applying various techniques such as group by, joins, subqueries, and window functions to uncover insights on property prices, trends, and geographical distribution.
- Created an interactive Tableau dashboard to visualize findings, providing a clear understanding of market dynamics and supporting data-driven decision-making.
- Analyzed a massive dataset of 5 million records for a bike-sharing company, utilizing a structured analytical approach. That included formulating questions, preparing and processing the data, performing extensive analysis, and effectively communicating findings to support informed business decision-making.
- The project uncovered valuable insights and actionable recommendations to improve customer experience, boost profitability, and drive business growth.
- Performed comprehensive exploratory data analysis on insurance charges, examining variable distributions and correlations to identify key cost drivers and their impacts.
- Developed a robust linear regression model with an accuracy of 86% in predicting insurance charges, supporting data-driven decision-making and cost optimization strategies for insurance providers.
