Soroush H. Zargarbashi
Scholar

Soroush H. Zargarbashi

Google Scholar ID: XnPiy-YAAAAJ
CISPA Helmhotz Center for Information Security
Trustworthy Machine LearningConformal PredictionUncertainty QuantificationGNNs
Citations & Impact
All-time
Citations
136
 
H-index
4
 
i10-index
4
 
Publications
8
 
Co-authors
0
 
Resume (English only)
Academic Achievements
  • 1. Optimal Conformal Prediction under Epistemic Uncertainty (Under review)
  • Co-authors: Alireza Javanmardi, Soroush H. Zargarbashi, Aleksandar Bojchevski, Willem Waegeman, Eyke Hüllermeier
  • 2. Conformal Prediction Sets for Graph Neural Networks (ICLR 2024)
  • Co-authors: Soroush H. Zargarbashi, Simone Antonelli, Aleksandar Bojchevski
  • 3. Robust Conformal Prediction with a Single Binary Certificate (Under review)
  • Co-authors: Soroush H. Zargarbashi, Aleksandar Bojchevski
  • 4. One Sample is Enough to make Conformal Prediction Robust (Under review)
  • Co-authors: Soroush H. Zargarbashi, Sadegh Akhondzadeh, Aleksandar Bojchevski
Research Experience
  • Research Assistant at CISPA Helmholtz Center for Information Security
  • Time: 2022-Ongoing
  • Work: Research on Trustworthy AI, Uncertainty Quantification, Conformal Prediction, Adversarial Robustness of machine learning models.
  • Research Intern at Apple
  • Time: May-Oct 2025
  • Work: Researching on speech-based translation systems.
Education
  • PhD Student: University of Cologne
  • Time: 2023-Ongoing
  • Supervised By: Prof. Dr. Aleksandar Bojchevski
Background
  • Research Interests: Uncertainty Quantification, Robustness of Machine Learning Models, Trustworthy AI. Works extensively with conformal prediction methods to provide reliable uncertainty estimates. Has hands-on experience in fine-tuning large language models for speech translation.
Co-authors
0 total
Co-authors: 0 (list not available)