Root Cause
Root cause analysis (RCA) aims to identify the underlying causes of system failures or anomalies across diverse domains, from autonomous driving to manufacturing and healthcare. Current research heavily emphasizes data-driven approaches, employing machine learning models like graph neural networks, transformers, and causal discovery algorithms (including Bayesian networks) to analyze complex datasets (often multimodal) and infer causal relationships. These advancements are improving the accuracy and efficiency of RCA, leading to faster troubleshooting, enhanced system reliability, and more informed decision-making in various industries. The development of large, publicly available datasets is also a significant focus, facilitating more robust benchmarking and algorithm comparison.
Papers
BALANCE: Bayesian Linear Attribution for Root Cause Localization
Chaoyu Chen, Hang Yu, Zhichao Lei, Jianguo Li, Shaokang Ren, Tingkai Zhang, Silin Hu, Jianchao Wang, Wenhui Shi
Supporting Safety Analysis of Image-processing DNNs through Clustering-based Approaches
Mohammed Oualid Attaoui, Hazem Fahmy, Fabrizio Pastore, Lionel Briand