Summary of Federated Split Learning For Human Activity Recognition with Differential Privacy, by Josue Ndeko et al.
Federated Split Learning for Human Activity Recognition with Differential Privacy
by Josue Ndeko, Shaba Shaon, Aubrey Beal, Avimanyu Sahoo, Dinh C. Nguyen
First submitted to arxiv on: 9 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
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 Federated Split Learning (FSL) with Differential Privacy (DP) for intelligent human activity recognition (HAR) over edge networks. The FSL-DP framework combines accelerometer and gyroscope data, leading to improved HAR accuracy. A comparison between traditional Federated Learning (FL) and FSL demonstrates the superiority of FSL in terms of accuracy and loss metrics. The study also explores the privacy-performance trade-off under different data settings, highlighting the balance between privacy guarantees and model accuracy. Moreover, the results show that FSL achieves faster communication times per training round compared to traditional FL, emphasizing its efficiency and effectiveness. This work contributes valuable insights and a novel framework tested on a real-life dataset. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research paper is about using computers to recognize human activities like walking or running. They created a new way for computers to learn from people’s movements called Federated Split Learning (FSL). The FSL method uses data from both accelerometers and gyroscopes, which are devices that track movement. This helps the computer recognize activities more accurately. The researchers also compared their method with another popular approach called Federated Learning (FL) and found that FSL is better in many ways. Additionally, they looked at how well their method keeps personal information private while still being accurate. Overall, this study shows a new way to recognize human activities using computers. |
Keywords
* Artificial intelligence * Activity recognition * Federated learning
