Statistics Student at Marmara University | TÜBİTAK STAR Scholarship Researcher

As a Statistics student, I am deeply fascinated by the mathematical foundations of models and the "stories" hidden in the details of data. I don't just build models; I strive to understand the underlying distributions and anomalies that others might overlook.

My approach to Data Science is Engineering-First: I bridge the gap between statistical rigor and scalable software architecture, ensuring that every piece of code is modular, maintainable, and production-ready.

• Statistical Modeling: Passionate about deep-diving into model assumptions, ensemble learning, and fuzzy logic systems.
• Architectural Mindset: Applying software engineering principles (like N-Tier Architecture and Clean Code) to data science workflows.
• Industrial AI: Focused on human-in-the-loop systems and turning "imperfect" real-world sensor data into actionable insights.
• MLOps & Reproducibility: Building automated, orchestrated pipelines (Airflow) to move models from notebooks to production.

• Languages: Python (Pandas, Scikit-learn, CatBoost), R (Tidyverse, Package Development), SQL.
• Engineering Foundation: Background in C# and N-Tier architecture, applied to modular data system design.
• Tools & Workflows: Apache Airflow, Git, LaTeX, Docker (learning).

• MFF (Meta Fuzzy Function): An R package developed under academic supervision for the TÜBİTAK STAR program. The project focuses on integrating ensemble learning with fuzzy clustering-based meta-modeling to enhance predictive performance.
• End-to-End Vehicle Pipeline: A robust, Airflow-orchestrated regression pipeline designed for vehicle price estimation. It features automated data transformation and utilizes prediction errors as a diagnostic signal for identifying real-world data anomalies.
• Synthetic Data & Regression Analysis: A comprehensive statistical study involving the generation of synthetic datasets based on scraped automotive data. This project focuses on evaluating regression assumptions, handling non-normal distributions, and performing advanced model diagnostics.

• LinkedIn: [https://www.linkedin.com/in/sad%C4%B1k-%C3%A7oban-5239aa253]
• Email: [s.c_2004@hotmail.com]
• Location: Istanbul, Turkey