Fault Localisation

Fault localization aims to identify the root causes of errors in complex systems, a crucial step in debugging and improving reliability. Current research emphasizes leveraging causal inference and dataflow graphs to pinpoint faulty components, particularly in data-driven applications and deep learning models. Techniques range from autoencoder-based anomaly detection with explained fault localization in industrial systems to learning-based approaches that directly identify faulty code segments in deep learning programs. These advancements improve software and system dependability, leading to more robust and reliable applications across various domains.