Publication
Explainable Latent-Space and Anomaly-Sensitive Learning for Thermodynamic Crystal Stability Prediction
2026
Within the scope of this study, on a large dataset consisting of 10,000 crystal materials; We have developed and tested a multimodal and ex…
Publication
Investigating Tree-Based Models Across Positive–Unlabeled Learning Frameworks for Crystal Synthesizability Prediction
2026
The majority of crystal structures in materials databases (Materials Project, etc.) are theoretically generated and have not been experimen…
Publication
Hybrid Graph–Machine Learning Framework for Accurate and Interpretable Band Gap Prediction
2026
Prediction of the electronic band gap, a critical parameter in semiconductor and energy materials, was carried out by combining various mac…
Project
Tubitak 2209a - Multi-Channel Image-Based Hybrid Model for Stellar Spectral Classification
2026
A hybrid AI architecture developed using multi-band (U, G, R, I, Z) photometric data and RGB images from the SDSS (Sloan Digital Sky Survey…