Berke Akkaya

Researcher in Machine Learning & AI | Ph.D. in Quantitative Methods, Istanbul University

View My GitHub Profile

Berke Akkaya

Projects • Publications • Teaching Activities • Contact

👋 Hello

I am a researcher working on recommender systems, natural language processing, and the empirical evaluation of machine learning models. I completed my PhD in Quantitative Methods at Istanbul University in 2026, with a thesis on content-aware product recommendation using sentence-transformer representations. During 2024–2025 I spent a year as a doctoral exchange researcher at the Chair of Data and Knowledge Engineering, University of Passau, working on large language model based recommendation.

My current interests centre on how models behave once they meet real users: how evaluation choices affect which model appears to perform best, and how performance varies across users and datasets.

Interests

Education

Publications

📚 Transparency in Candidate-Facing Recruitment Processes in the BIST 30 Context

25th International Business Congress, Ankara, pp. 1–13 (2026)
→ With E. Çelik. This paper examines how automated and algorithmic steps in hiring are disclosed to candidates..

📚 Current Trends in Recommender Systems: A Survey of Approaches and Future Directions

Computer Science, 10(1), 53-91 (2025)
→ This paper discusses the growing importance of recommender systems in enhancing user experience and information access in digital environments.
🔗 Paper Link

📚 A Comparative Study of Heart Disease Prediction Using Machine Learning Techniques

2022 International Congress on Human-Computer Interaction, Optimization and Robotic Applications (HORA), IEEE
→ Compared ML models on CDC BRFSS 2020 data; achieved 0.89 accuracy with XGBoost after SMOTE-Tomek balancing.
🔗 Paper Link

📚 Global Vaccination Inequality and Influencing Factors During COVID-19

Hacettepe Journal of Health Administration, 25(1), 17-28 (2022)
→ Analyzed political, economic, and demographic drivers of vaccine access across 88 countries using R and SPSS. 🔗 Paper Link

📚 The Mediator Role of Economic Freedom in the Effect of Corruption Perception on National Happiness

Izmir Journal of Economics, 37(3), 760–777 (2022)
→ Explored how economic freedom influences the link between perceived corruption and national happiness in 150 countries. 🔗 Paper Link

📚 The Effect of Recursive Feature Elimination with Cross-Validation on Classification Performance

4th International Conference on Data Science and Applications, 142–145 (2021)
→ Tested recursive feature elimination on 9 ML algorithms across 5 datasets. Improved accuracy and reduced fit time. 🔗 Paper Link

📚 Comparison of Multi-Class Classification Algorithms for Early Heart Disease Diagnosis

Y-BIS 2019 Conference, ISBIS Young Statisticians Workshop
→ Compared 9 classification algorithms on UCI cardiotocography data. Best model: Random Forest (0.88 accuracy). 🔗 Paper Link

Projects

🎵 Spotify Recommendation System

A content-based recommendation system that uses cosine similarity on standardized audio features to suggest similar songs.
Tools: Python, Spotify API, Cosine Similarity
📂 Tags: Recommender, Content-Based, Audio Analysis

👗 H&M Fashion Recommendation

Collaborative filtering recommendation system using transaction data and language models.
Tools: Python, NLP, Transaction Logs
📂 Tags: Fashion, Collaborative Filtering, Language Models

📖 Sentiment Analysis of Poe’s Corpus

NLP-based sentiment analysis on Edgar Allan Poe’s short stories using VADER and visualized emotion trends.
Tools: Python, VADER, Word Clouds
📂 Tags: NLP, Sentiment, Visualization

🏥 Heart Disease Prediction with ML

Comparative study on 9 ML models for early heart disease prediction using SMOTE-Tomek and CDC BRFSS dataset.
Tools: Python, XGBoost, Sklearn
📂 Tags: Health, Classification, SMOTE

Teaching

Basic Mathematics I & II

Fundamentals of algebra, functions, and calculus applied to economics and business problems.

Statistical Analysis

Practical implementation of hypothesis testing, confidence intervals, and regression analysis.

Operations Research

Application of linear programming and optimization techniques in decision-making scenarios.

Introduction to Statistics

Basic descriptive statistics and probability theory for beginners.

Business Mathematics

Quantitative methods used in finance and business management, including interest calculations.

Statistics with R

Hands-on use of the R programming language for data analysis and statistical modeling.

Data Mining

Introduction to data preprocessing, clustering, classification, and association rule learning.

CV

📄 You can download my latest resume here.

Contact

📬