Skip to content
View peyamikenanoglu's full-sized avatar

Block or report peyamikenanoglu

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
peyamikenanoglu/README.md

Hi, I'm Peyami Kenanoğlu

AI Engineer & Data Scientist focused on end-to-end machine learning systems, applied AI, and portfolio-grade data products.

LinkedInORCIDGitHub


About Me

I build practical machine learning and AI projects with an emphasis on real-world validation, leakage-safe modeling, feature engineering, and clear project documentation.

My current focus is on building a strong AI/Data Science portfolio through end-to-end projects in:

  • Fraud detection
  • Customer churn prediction
  • Forecasting
  • Recommender systems
  • Applied machine learning pipelines
  • Deployment-oriented ML systems

I also have an academic background in Project and Construction Management, with PhD-level research on machine learning applications for construction risk assessment.


Featured Projects

IEEE-CIS Fraud Detection

Time-aware fraud detection pipeline using advanced feature engineering, UID-style features, LightGBM, XGBoost, CatBoost, and ensemble learning.

Tech: Python, pandas, scikit-learn, XGBoost, LightGBM, CatBoost
Focus: Fraud detection, feature engineering, leakage-safe validation, imbalanced classification
Result: ROC-AUC = 0.9292
Repository: peyamikenanoglu/ieee-fraud-detection


KKBox Churn Prediction

End-to-end customer churn prediction project using the WSDM KKBox dataset, with realistic time-aware validation instead of an overly optimistic random split.

Tech: Python, pandas, scikit-learn, XGBoost
Focus: Churn prediction, temporal validation, customer behavior modeling
Repository: peyamikenanoglu/kkbox-churn-prediction


TMDB Box Office Prediction & Movie Recommender

Machine learning project for box office revenue prediction and movie recommendation using structured movie metadata.

Tech: Python, pandas, scikit-learn, gradient boosting models
Focus: Regression, recommender systems, feature engineering, model evaluation
Repository: peyamikenanoglu/tmdb-box-office-ml-portfolio


Technical Stack

Languages & Core Tools

Python SQL Git GitHub

Data Science & Machine Learning

pandas NumPy scikit-learn XGBoost LightGBM CatBoost

Modeling Areas

Classification Regression Forecasting Churn Prediction Fraud Detection Recommender Systems

Engineering & Deployment

FastAPI Docker Streamlit

Model Interpretation

SHAP Permutation Importance Feature Importance


Currently Working On

  • Strengthening my GitHub portfolio with business-oriented machine learning projects
  • Improving deployment skills with FastAPI, Docker, and Streamlit
  • Building clean, reproducible ML pipelines suitable for professional review
  • Expanding SQL and data engineering skills for AI/ML roles

Profile Direction

My goal is to build practical AI systems that go beyond notebooks: structured repositories, reproducible pipelines, readable documentation, and deployable machine learning products.

Pinned Loading

  1. ieee-fraud-detection ieee-fraud-detection Public

    End-to-end fraud detection system with time-aware validation, advanced feature engineering, LightGBM/XGBoost/CatBoost ensemble, FastAPI prediction service, and Streamlit frontend.

    Python

  2. kkbox-churn-prediction kkbox-churn-prediction Public

    End-to-end churn prediction project using the KKBox dataset

    Python

  3. tmdb-box-office-ml-portfolio tmdb-box-office-ml-portfolio Public

    End-to-end box office prediction and movie recommender in Python

    Python