Articles

Intelligent but Opaque: The Black Box of Artificial Intelligence and the Explainability Revolution

Why is it important to understand the decisions made by artificial intelligence? In this article, Dr. Ozan Önal explores the AI black-box problem, Explainable Artificial Intelligence (XAI), and methods such as SHAP and LIME. Using real-world examples, he discusses the relationship between explainability, trustworthiness, fairness, accountability, and human oversight.

When We Pour Our Hearts Out to Artificial Intelligence…

A column examining the privacy and security implications of sharing personal information with generative artificial intelligence systems, from private conversations and health information to photographs and documents.

The Pioneering Role of Physicists in Artificial Intelligence

Published in Issue 83 of the FMO e-Bulletin, this article by Dr. Ozan Önal explores physicists’ contributions to artificial intelligence and the relationship between physics and machine learning.

The Bottomless Pit of Artificial Intelligence

“The Bottomless Pit of Artificial Intelligence” is an article written by Dr. Ozan Önal and published in Issue 196 of Bilişim Kültürü Dergisi by the Informatics Association of Türkiye. The article introduces fundamental concepts of artificial intelligence and machine learning, including supervised and unsupervised learning, classification, regression, data security, algorithmic reliability, ethics, and social responsibility. It also emphasizes the importance of developing trustworthy, fair, transparent, and responsible artificial intelligence systems.

A Photograph Is More Than Just a Photograph

A media feature including Dr. Ozan Önal’s views on AI-powered photo applications, facial data processing, biometric data, digital privacy and trustworthy artificial intelligence.

The effect of Ti on Mechanical, Morphological and Magnetic Properties in Fe-Ni-Ti Alloys

Morphological, mechanical and magnetic properties of Fe-Ni-Ti alloys with different Ti contents were investigated using SEM, Mössbauer spectroscopy, DSC and Vickers hardness measurements.

Prediction of Ionic Conductivity in LLZO-Based Solid Electrolytes Using Machine Learning and Determination of Optimum Compositions

This study aims to predict the ionic conductivity of LLZO (Li₇La₃Zr₂O₁₂)-based solid electrolytes using machine learning methods. A comprehensive preprocessing pipeline was applied to a literature-based dataset, multiple machine learning models were compared, and the best overall performance was achieved through ensemble learning approaches. After outlier analysis, the model performance reached approximately R² ≈ 0.89. This work was presented at the 3rd Bilsel International Erzurum Scientific Research Congress and published in the congress book.

Dr. Önal: Artificial Intelligence Has Become One of the Strategic Fields Directly Shaping the Future of Nations

AI specialist Dr. Ozan Önal shared his views on his doctoral research, the future of artificial intelligence and data science, explainable and trustworthy AI, data governance, and standardization.

Effect of Ti Element on Martensitic Transformations in Fe-Ni-x%Ti Alloys

An academic study investigating the effect of Ti addition on martensitic transformations in Fe-Ni-x%Ti alloys.