I enjoy combining my gustos with data. Holding MSc. Applied Data Science at Utrecht University, the Netherlands.

• My latest personal project: "Shame Network". I developed a website called rezilsiniz.com, targeting Turkish celebrities, influencers, and public figures who remained silent during a politically charged boycott movement. This movement, sparked by the arrest of Ekrem İmamoğlu in March 2025, escalated into a nationwide anti-government protest. The project was developed to reflect public frustration over their lack of solidarity.

  • I built the website using TypeScript and JavaScript. I know it's not visually stunning yet. I'm still learning and improving my skills in these languages. But I'm hopeful it will get better with time!

• 🌱 On the way of learning Docker

• Programming languages :
  

• Machine Learning / Deep Learning / Data Analysis frameworks :
  

• Other :
  

Applying data science and machine learning techniques to financial datasets for analysis, prediction, and optimization.

• Stock Market Analysis - StockVision is an interactive dashboard that provides comprehensive stock market data visualization and analysis, including advanced price charts, technical indicators, volume tracking, and AI-powered forecasting models like moving averages, linear regression, and ARIMA, all presented in a clean, user-friendly interface.
• GDP Emission Dashboard - A comprehensive data visualization and analysis dashboard that explores the relationships between global GDP, population, and CO2 emissions from 1977 to 2022. The dashboard includes interactive visualizations, historical analysis, and a machine learning model for CO2 emissions prediction.
• Fraud Detection Pipeline - A comprehensive fraud detection pipeline that uses advanced machine learning techniques to identify fraudulent financial transactions. It includes smart data loading, feature engineering, sophisticated model ensembling, and comprehensive visualizations to achieve high-performance fraud detection

Projects focused on the operationalization of machine learning models, including deployment, monitoring, and maintenance.

• Diabetes Prediction API - Predicting diabetes using the Pima Indians Diabetes dataset with R and Plumber API for deployment.
• Automated ML Pipeline - An automated ML pipeline which trains different models, tracks experiments using MLflow, serves predictions through a FastAPI endpoint, and handles the entire ML lifecycle in a structured way.
• Palmer Penguins MLOps Project - A complete MLOps pipeline in R, using the Palmer Penguins dataset to predict penguin sex based on various physical measurements. It showcases modern machine learning operations practices including model development, hyperparameter tuning, versioning, and deployment.

Machine learning applications and models developed to solve various real-world problems and datasets.

• Hyperparameter Optimization Comparison - This project demonstrates different hyperparameter optimization techniques using scikit-learn's breast cancer dataset. It compares three different optimization methods: Grid Search CV, Random Search CV, and Optuna Optimization.
• Object Detection with YOLOv3 - This project uses YOLOv3, a state-of-the-art object detection algorithm, to identify and locate objects in images.

Implementing machine learning algorithms and statistical methods from scratch to deepen understanding of core concepts.

• Principal Component Analysis (PCA) from Scratch - This repository contains a comprehensive implementation of Principal Component Analysis (PCA) built from scratch in Python. The implementation includes various scaling options, variance explanation, data projection and reconstruction, outlier detection, and multiple visualization techniques.
• Neural Network from Scratch - This project implements a neural network from scratch using NumPy, with a focus on the MNIST dataset. It includes features like Adam optimizer, Categorical Cross Entropy loss function, mini-batch gradient descent, and comprehensive model evaluation visualizations.