Understanding the Bug Characteristics and Fix Strategies of Federated Learning Systems
Federated learning (FL) is an emerging machine learning paradigm that aims to address the problem of isolated data islands. To preserve privacy, FL allows machine learning models and deep neural networks to be trained from decentralized data kept privately at individual devices. FL has been increasingly adopted in mission-critical fields such as finance and healthcare. However, bugs in FL systems are inevitable and may result in catastrophic consequences such as financial loss, inappropriate medical decision, and violation of data privacy ordinance. While many recent studies were conducted to understand the bugs in machine learning systems, there is no existing study to characterize the bugs arising from the unique nature of FL systems. To fill the gap, we collected 395 real bugs from six popular FL frameworks (Tensorflow Federated, PySyft, FATE, Flower, PaddleFL and Fedlearner) in GitHub and StackOverflow, and then manually analyzed their symptoms and impacts, prone stages, root causes and fix strategies, and report a series of findings and actionable implications. Finally, we provide possible suggestions or solutions for developers of FL systems based on the above findings and implications.
Thu 7 DecDisplayed time zone: Pacific Time (US & Canada) change
11:00 - 12:30 | Machine Learning IVResearch Papers / Ideas, Visions and Reflections / Industry Papers at Golden Gate C2 Chair(s): Diptikalyan Saha IBM Research India | ||
11:00 15mTalk | Dynamic Data Fault Localization for Deep Neural Networks Research Papers Yining Yin Nanjing University, China, Yang Feng Nanjing University, Shihao Weng Nanjing University, Zixi Liu Nanjing University, Yuan Yao Nanjing University, Yichi Zhang Nanjing University, Zhihong Zhao , Zhenyu Chen Nanjing University Media Attached | ||
11:15 15mTalk | Assisting Static Analysis with Large Language Models: A ChatGPT Experiment Ideas, Visions and Reflections Haonan Li University of California at Riverside, USA, Yu Hao University of California at Riverside, USA, Yizhuo Zhai University of California at Riverside, USA, Zhiyun Qian University of California at Riverside, USA Media Attached | ||
11:30 15mTalk | Understanding the Bug Characteristics and Fix Strategies of Federated Learning Systems Research Papers Xiaohu Du Huazhong University of Science and Technology, Xiao CHEN Department of Computer Science and Engineering, The Hong Kong University of Science and Technology, Jialun Cao Hong Kong University of Science and Technology, Ming Wen Huazhong University of Science and Technology, Shing-Chi Cheung Department of Computer Science and Engineering, The Hong Kong University of Science and Technology, Hai Jin Huazhong University of Science and Technology Media Attached | ||
11:45 15mTalk | EvoCLINICAL: Evolving Cyber-Cyber Digital Twin with Active Transfer Learning for Automated Cancer Registry System Industry Papers Chengjie Lu Simula Research Laboratory; University of Oslo, Xu Qinghua Simula Research Laboratory; University of Oslo, Tao Yue Beihang University, Shaukat Ali Simula Research Laboratory and Oslo Metropolitan University, Thomas Schwitalla Cancer Registry of Norway, Jan F. Nygård Cancer Registry of Norway DOI Media Attached | ||
12:00 15mTalk | Learning Program Semantics for Vulnerability Detection via Vulnerability-specific Inter-procedural Slicing Research Papers bozhi wu Singapore Management University, Shangqing Liu Nanyang Technological University, Yang Xiao Institute of Information Engineering at Chinese Academy of Sciences; University of Chinese Academy of Sciences, Zhiming Li Nanyang Technological University, Singapore, Jun Sun Singapore Management University, Shang-Wei Lin Nanyang Technological University Media Attached | ||
12:15 15mTalk | [Remote] DeepRover: A Query-efficient Blackbox Attack for Deep Neural Networks Research Papers Fuyuan Zhang Kyushu University, Xinwen Hu Hunan Normal University, Lei Ma The University of Tokyo / University of Alberta, Jianjun Zhao Kyushu University Media Attached | ||