In high-energy physics experiments, especially in the CMS and ATLAS experiments at the Large Hadron Collider (LHC) at CERN, millions of collisions take place every second, generating petabytes of data. Processing, analyzing and transforming this data into meaningful physical results has become too complex to be limited by traditional analysis methods. Therefore, in recent years, machine learning (ML) and deep learning (DL) techniques have been widely used in this field.
In this study, we focus on ML-based methods developed on particle physics data. Various unsupervised and semi-supervised learning approaches are applied using large volumes of simulated data derived from Standard Model processes. Variational autoencoder, normalizing flow and graph neural network architectures were preferred for the analysis of jet structures. Through these models, anomaly detection was performed and the discriminability of events potentially carrying New Physics signatures was evaluated.
During the data preprocessing, model training and evaluation phases, care was taken to maintain physical consistency. In particular, domain adaptation techniques were utilized to reduce mismatches between simulation and real data, and explainable artificial intelligence (XAI) approaches were used to improve the interpretability of model results. Significant success has been achieved in tasks such as improving triggering decisions in experimental systems, detecting rare events and extracting physical patterns in high-dimensional data space.
In this presentation, we will share current approaches to data science and deep learning methods applied on data from LHC experiments and show how they contribute to the search for new physics in particle physics. Furthermore, it will be emphasized that the methods presented have the potential to be applicable not only for fundamental sciences but also for other fields that require smart data analytics.

Prof. Dr. Bora Işıldak works on the search for new physics in high-energy particle collisions, jet physics and data science applications. He completed his Ph.D. at Boğaziçi University, where he took an active role in the CMS (Compact Muon Solenoid) experiment at CERN. He is currently a professor in the Department of Physics at Yıldız Technical University and continues his research in the CMS collaboration.
His research interests include the search for exotic particles, quantum chromodynamics (QCD), high-dimensional data analysis and the applications of artificial intelligence methods to particle physics. In recent years, he has been working on the development of anomaly detection, rare event discrimination and triggering systems using modern deep learning approaches such as graph neural networks, variational automodelers and normalizing flows.
He utilizes domain adaptation techniques to eliminate the differences between experimental data and simulations, and explainable artificial intelligence (XAI) methods to increase the interpretability of model outputs. These studies have the potential to be applied not only in basic sciences but also in other disciplines that require intelligent data analysis.

The advent of generative artificial intelligence (AI) is catalyzing transformational change across all industrial sectors. While conventional machine‑learning applications have largely pursued engineering optimization, generative AI inaugurates a paradigm that privileges the synthesis of novel alternatives and design trajectories. This generative capacity is especially consequential for architecture and, more specifically, architectural design. In this keynote, I present a suite of AI‑BIM applications developed through a Korea Government-funded Project team’s research initiative, “Automation of AI‑Based Architectural Design.” By situating these tools within the broader, rapidly evolving generative‑AI landscape, the talk will demonstrate how such technologies can be strategically deployed to accelerate design workflows, expand creative exploration, and redefine professional practice in architecture.

Jin-Kook Lee is a Professor of Interior Architecture at Yonsei University, specializing in AI & design computing, Building Information Modeling (BIM), and digital design methods. His research focuses on spatial analysis automation, design visualization, and architectural planning & simulation. He has published extensively in international journals and conferences. His work has been recognized with the BIM Award from the AIA and the Technology Innovation Award from FIATECH. He has also received honors in international digital design competitions. His academic and industry projects emphasize the integration of computing technologies for creating better living environments.