Yong Xu
Senior Fellow, Microsoft Internet Engineering Asia
Senior Researcher at Microsoft Asia Internet Engineering Institute.He joined Microsoft after graduation in 2017, and has held several research positions in Microsoft Research Asia and Microsoft Asia Internet Engineering Institute. His research interests mainly involve the challenges of data analytics and machine learning algorithms in specific applications, especially the practical implementation in cloud service scenarios, such as anomaly detection, intelligent diagnosis, log analysis, etc. He has published several papers in the top conferences in software engineering and systems field USENIX ATC, ICSE, FSE, EuroSys, etc. Currently, he is also exploring the enhancement of cognitive reasoning ability of big models, and the efficient application of big models in vertical industrial scenarios.
Topic
Intelligent O&M - Data, Models and Applications
Introduction：This presentation will focus on the challenges in various scenarios of intelligent operation and maintenance (AIOps), and share our deep cultivation and innovation in technology. Specifically, in anomaly detection, automatic diagnosis, security deployment and other scenarios, how to analyze and mine large-scale system operation indexes and log data, and combine with the needs of actual operation and maintenance scenarios to put forward general and efficient solutions. For example, in anomaly detection, we propose a meta-heuristic search-based approach for generalized and efficient fault detection for massive high-dimensional data. To address the problems of high false alarm rate and low interpretability of complex deep learning methods in microservice scenarios, we further propose a performance anomaly detection method based on Trace analysis to improve accuracy and interpretability. We will also share the current attempts to apply big models on AIOps.
