An empirical study on TensorFlow program bugs
By Yuhao Zhang, Yifan Chen, Shing-Chi Cheung, Yingfei Xiong, Lu Zhang. International Symposium on Software Testing and Analysis, 2018
335 citations recorded by the source corpus.
Deep learning applications become increasingly popular in important domains such as self-driving systems and facial identity systems. Defective deep learning applications may lead to catastrophic consequences. Although recent research efforts were made on testing and debugging deep learning applications, the characteristics of deep learning defects have never been studied. To fill this gap, we studied deep learning applications built on top of TensorFlow and collected program bugs related to TensorFlow from StackOverflow QA pages and Github projects. We extracted information from QA pages, commit messages, pull request messages, and issue discussions to examine the root causes and symptoms of these bugs. We also studied the strategies deployed by TensorFlow users for bug detection and localization. These findings help researchers and TensorFlow users to gain a better understanding of coding defects in TensorFlow programs and point out a new direction for future research.
Subfields: Artificial Intelligence, Signal Processing
View Paper: https://www.semanticscholar.org/paper/aca824950a7f0c58b10c963fe2e16673b00e6f83
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Mark Harman
- Paolo Tonella
- Gabriele Bavota
- Lei Ma
- Yang Liu
- Jie M. Zhang
- Felix Stahlberg
- Graham Neubig
In Signal Processing
Citations in the corpus
- Machine Learning Testing: Survey, Landscapes and Horizons (2019, background): “[160] studied 175 Tensorflow bug reports from StackOverflow QA (Question and Answer) pages and from Github projects.”
- Taxonomy of Real Faults in Deep Learning Systems (2019, methodology): “To ease the comprehension, we list the categories with the exact naming and excerpts of descriptions from the corresponding publication [30]: 1.”
- A comprehensive study on deep learning bug characteristics (2019, methodology): “We have adapted the classification scheme of root causes and bug effects from [25] and added on top of that as found from the study of the posts.”
- Software Engineering for AI-Based Systems: A Survey (2021, background): “Three of these papers conducted empirical studies examining bugs in ML projects [91, 190, 229], whereas one paper proposed a specialized debugger for ML models [34].”