![]() Furthermore, the patterns related to subprocess Python basic syntax, model training, parallelization, data transformation, and Low-level types from the stack trace patterns: most patterns are related to Third, we derive five high-level categories and 25 ![]() Traces, even across different ML libraries, with a small portion of patternsĬovering many stack traces. Second, we observe that recurrent patterns exists in ML stack Than questions without stack traces however, they are less likely to getĪccepted answers. We observe that ML questions that contain stack traces gain more popularity Study 11,449 stack traces related to seven popular Python ML libraries. To that end, we mine Stack Overflow (SO) and Thus, studying the patterns in stack traces can help practitionersĪnd researchers understand the causes of exceptions in ML applications and theĬhallenges faced by ML developers. Indeed, theseĮxceptions may cross the entire software stack (including applications and These stack traces describe the chain ofįunction calls that lead to an anomalous situation, or exception. Explicit programming errors usually manifest throughĮrror messages and stack traces. ![]() Traditional software, ML applications are not immune to the bugs that resultįrom programming errors. Tremendous popularity in a wide range of applications. Download a PDF of the paper titled What Causes Exceptions in Machine Learning Applications? Mining Machine Learning-Related Stack Traces on Stack Overflow, by Amin Ghadesi and 2 other authors Download PDF Abstract: Machine learning (ML), including deep learning, has recently gained
0 Comments
Leave a Reply. |