Research
Research
Overview
Information and Learning Theory Lab. (ILT Lab.) at DGIST, led by Daewon Seo, explores a broad spectrum of topics in wireless communications and machine learning from a theoretical standpoint. Specifically, our focus is on data communications (5G/6G, storage systems) and inference natures in machine learning and social networks, examined through the perspectives of statistical inference, information theory, and learning theory.
(Beyond) 6G-targeted wireless communications
6G and beyond 6G wireless communication systems are expected to incorporate a range of emerging technologies—such as ISAC (integrated sensing and communication), semantic communication, and joint source-channel coding—to overcome the limitations of current systems. We investigate these enabling technologies from both theoretical and practical perspectives, including implementations based on deep learning.
Fundamental limits of communication and storage systems
Noise is inherent in our information processing systems such as communication, storage, and learning systems, which means that our system may frequently malfunction. Therefore, it is crucial to know how to configure systems under such noise, what is the fundamental performance limits, and how to attain these limits. We study these issues from an information-theoretic perspective. Typical topics and target applications include wireless communications (e.g., 5G, 6G) and data storage systems (e.g., JPEG, MPEG).
Efficient/scalable ML
The rapid development of data collection methods enables us to obtain large volumes of high-dimensional data. Moreover, the success of overparameterized neural networks requires ML algorithms to be efficient and scalable. There are several approaches addressing this challenge, such as implementing ML algorithms in a distributed manner and pre-/post-processing data for better inference. We explore this topic from a theoretical perspective.