Skip to main content

Seminars

  • PRIVACY-PRESERVING ANOMALY DETECTION FOR SMART ENERGY MANAGEMENT

    PhD Thesis📅 27.02.2026 — 10:30 👤 Speaker: MANOLYA ATALAY 🎓 Supervisor(s): ASSOC.PROF.DR.PELIN ANGIN ULKUER 📍 Location: A105 ⏲ Duration: 120 min. 📝 Abstract: The widespread deployment of smart meters in modern energy grids introduces not only unprecedented opportunities for high-frequency analytics, but also acute privacy and…

  • Natural Language Interfaces for Databases: Approaches to the Problem of Query Generation

    PhD Thesis📅 23.02.2026 — 14:30 👤 Speaker: ARIF GORKEM OZER 🎓 Supervisor(s): PROF.DR.I.HAKKI TOROSLU,PROF.DR.PINAR KARAGOZ 📍 Location: BMB5 ⏲ Duration: 120 min. 📝 Abstract: The number of databases, as well as their size and complexity, is steadily increasing. This trend creates a significant barrier to data access, especially…

  • Enhancing Efficiency in Large-Scale Search: From Traditional to Neural Retrieval Systems

    PhD Thesis📅 23.02.2026 — 13:30 👤 Speaker: ERMAN YAFAY 🎓 Supervisor(s): PROF.DR.ISMAIL SENGOR ALTINGOVDE 📍 Location: A105 ⏲ Duration: 120 min. 📝 Abstract: Efficient retrieval of relevant documents from massive collections remains an essential challenge in Information Retrieval (IR). Modern search engines face immense…

  • Advancing Dialogue Systems with Temporal Contextual Topic Forecasting and Controlled Response Retrieval

    PhD Thesis📅 23.02.2026 — 13:00 👤 Speaker: EMRE KULAH 🎓 Supervisor(s): ASSOC.PROF.DR.HANDE ALEMDAR 📍 Location: A101 ⏲ Duration: 120 min. 📝 Abstract: Dialogue systems powered by large language models (LLMs) show strong generative abilities but often struggle with informal language, long-term coherence, and grounded responses in…

  • Improving the performance of prioritized planning in multi-agent path finding

    MSc Thesis📅 23.01.2026 — 13:30 👤 Speaker: ANIL CABAK 🎓 Supervisor(s): PROF.DR.FARUK POLAT 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: MAPF is a fundamental challenge that is frequently encountered in robotics, warehouse automation, and systems with more than one robot. The goal is to find paths for multiple agents that don…

  • FROM SYNTHETIC MEDIUM TO REAL-WORLD APPLICATION: FINE-TUNING A MEDICAL LLM FOR DDX

    MSc Thesis📅 23.01.2026 — 10:00 👤 Speaker: EZGI CAVAS 🎓 Supervisor(s): PROF.DR.NIHAN KESIM CICEKLI, DR.AYSENUR BIRTURK 📍 Location: A101 ⏲ Duration: 90 min. 📝 Abstract: Access to large-scale, annotated EHR is limited by privacy rules. This creates a major setback for training strong clinical NLP models. Synthetic data provides a…

  • GRAPH-BASED REPRESENTATION LEARNING FROM LARGE-SCALE BIOMEDICAL NETWORKS

    PhD Thesis📅 22.01.2026 — 13:00 👤 Speaker: GOKHAN OZSARI 🎓 Supervisor(s): PROF.DR.HALIT OGUZTUZUN,PROF.DR.M.VOLKAN ATALAY 📍 Location: A101 ⏲ Duration: 120 min. 📝 Abstract: Graph Neural Networks (GNNs) have gained significant attention in drug discovery due to their ability to model graph-structured biological and chemical data,…

  • SEMANTIC MATCHING FOR REVIEWER ASSIGNMENT: AN ABLATION STUDY OF DENSE RETRIEVAL AND LLM REASONING

    MSc Thesis📅 22.01.2026 — 10:00 👤 Speaker: IREM DERELI 🎓 Supervisor(s): PROF.DR.NIHAN KESIM CICEKLI 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: Assigning suitable academic reviewers to research projects is a complex task that requires understanding both the semantic content of project descriptions and the diverse research…

  • Surrogate Modeling of Aerodynamic Surfaces Using Geometric Deep Learning

    MSc Thesis📅 15.01.2026 — 11:00 👤 Speaker: HAKAN ARI 🎓 Supervisor(s): PROF.DR.MURAT MANGUOGLU, ASSOC.PROF.DR.HANDE ALEMDAR 📍 Location: A101 ⏲ Duration: 90 min. 📝 Abstract: Accurate calculation of physical quantities such as surface pressure, wall shear stress, and temperature is crucial in aerodynamic design processes. Data-…

  • Enhancing Retrieval Augmented Generation via Decomposional Reasoning and Active Learning

    MSc Thesis📅 13.01.2026 — 14:00 👤 Speaker: AHMET KAGAN KAYA 🎓 Supervisor(s): ASSOC.PROF.DR.R.GOKBERK CINBIS 📍 Location: A105 ⏲ Duration: 90 min. 📝 Abstract: The deployment of Large Language Models (LLMs) in knowledge-intensive domains is currently challenged by two critical bottlenecks: the computational cost of inference and…