Summary of Tpfl: a Trustworthy Personalized Federated Learning Framework Via Subjective Logic, by Jinqian Chen et al.
TPFL: A Trustworthy Personalized Federated Learning Framework via Subjective Logic
by Jinqian Chen, Jihua Zhu
First submitted to arxiv on: 16 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Distributed, Parallel, and Cluster Computing (cs.DC)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper introduces Trustworthy Personalized Federated Learning (TPFL), a framework designed for classification tasks via subjective logic. TPFL constructs federated models by employing subjective logic to provide probabilistic decisions with uncertainty assessments. It mitigates data heterogeneity by incorporating a trainable prior during local training and utilizes model and instance uncertainty to ensure reliable training and inference. The framework is demonstrated to achieve competitive performance on widely recognized benchmarks, while exhibiting resilience against attacks, robustness on domain shifts, and reliability in high-stake scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning helps computers learn together without sharing data. But it’s not always safe or fair. This paper tries to fix that by creating a new way to make sure the results are trustworthy. They use something called subjective logic to make predictions with uncertainty scores. This helps prevent mistakes and ensures the results are reliable. The method also helps deal with different types of data and is resistant to attacks. The authors tested their idea on well-known datasets and showed that it works well. |
Keywords
» Artificial intelligence » Classification » Federated learning » Inference
