Anomaly detection has been a lasting yet active research area for several decades in various research communities, including machine learning, data mining, computer vision and statistics. In recent years deep learning-enabled anomaly detection has demonstrated tremendous success in broad real-world applications, ranging from cyber-attack detection, fraud detection, crime detection, defect detection in our daily life to scientific discovery contexts like rover-based planetary exploration. This talk will first introduce some largely unsolved challenges in anomaly detection, followed by an overview of the speaker research in tackling these challenges and its exciting interdisciplinary application exemplars. The talk will then focus on how to address those challenges with some emerging deep anomaly detection techniques.
About the Speaker
Guansong Pang is a Research Fellow at the Australian Institute for Machine Learning, University of Adelaide, Australia, where he joined in September 2018. He received a PhD degree from University of Technology Sydney (UTS) in May 2019, wining the university-wide best doctorate dissertation award – UTS Chancellor’s Award List. His research interests are to explore elegant data mining and machine learning algorithms for abnormality and rarity detection, and to tackle real-world problems in healthcare, information security, safety in AI systems and physical worlds, and scientific discovery. He publishes regularly in top AI and Data Science conferences and journals such as KDD, CVPR, AAAI, IJCAI, ICDM, ACM MM, CIKM, ACM Computing Surveys, IEEE TKDE, ACM TKDD, DMKD (Springer), JAIR, IP&M, Bioinformatics, IEEE TMI, etc. He serves as (senior) PC member of these prestigious conferences and reviewer of the top journals. He is a (leading) guest editor of Special Issues with IEEE Transactions on Neural Networks and Learning Systems on deep anomaly detection and IEEE Intelligent Systems on non-i.i.d. anomaly detection. He is the organization chair of the IJCAI’20 Workshop on Artificial Intelligence for Anomalies and Novelties.
He is a tenure-track faculty candidate for the Artificial Intelligence & Data Science, Machine Learning & Intelligence cluster.
