[Seminar] Fast Outlier Detection from Data Streams
호스트: 강유 교수 (x7254)
Recently, the advancement of network technologies for cloud computing, IoT, and 5G is combined with the advancement of hardware technologies for semiconductors and sensors, and they facilitate the collection, management, and processing of data streams generated in real time. As a result, there has been growing needs for rapid acquisition of valuable information from real-time data streams in various industries. In particular, outlier detection techniques for finding abnormal data points that deviate significantly from normal data points are widely studied in many applications such as finance, manufacturing, and healthcare. In this talk, I present our recent work that aims to detect various types of outliers with high accuracy and low latency, mainly by preventing redundant updates of existing algorithms. This work has been presented at VLDB19, KDD20, and SIGMOD21.
온라인 줌 링크: https://snu-ac-kr.zoom.us/j/3349805948
Jae-Gil Lee is an Associate Professor at School of Computing, Korea Advanced Institute of Science and Technology (KAIST) and is leading Data Mining Lab. Before that, he was a Postdoctoral Researcher at IBM Almaden Research Center and a Postdoc Research Associate at Department of Computer Science, University of Illinois at Urbana-Champaign. At IBM, he was one of the key contributors to IBM Smart Analytics Optimizer. At University of Illinois, he worked on spatio-temporal data mining with Prof. Jiawei Han. He earned the Ph.D degree under the supervision of Prof. Kyu-Young Whang from KAIST in 2005. His research interests encompass mobility and stream data mining, deep learning-based big data analysis, and distributed deep learning. His lab has been selected as a SW Star Lab by the Ministry of Science and ICT in 2020.