Summary of Learning to Better See the Unseen: Broad-deep Mixed Anti-forgetting Framework For Incremental Zero-shot Fault Diagnosis, by Jiancheng Zhao et al.
Learning to better see the unseen: Broad-Deep Mixed Anti-Forgetting Framework for Incremental Zero-Shot Fault Diagnosis
by Jiancheng Zhao, Jiaqi Yue, Chunhui Zhao
First submitted to arxiv on: 18 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 Zero-shot fault diagnosis (ZSFD) enables the identification of unseen faults by predicting fault attributes labeled by human experts. This paper addresses the need for ZSFD to adapt to new fault categories and attributes while avoiding forgetting previously learned diagnoses in industrial processes. To overcome this challenge, the incremental ZSFD (IZSFD) paradigm is proposed, incorporating category increment and attribute increment for traditional ZSFD and generalized ZSFD. A broad-deep mixed anti-forgetting framework (BDMAFF) is presented to achieve IZSFD, learning from new fault categories and attributes while accumulating previously acquired knowledge through feature memory and attribute prototype memory. The effectiveness of BDMAFF is verified using a real hydraulic system and the Tennessee-Eastman benchmark process. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making machines that can diagnose problems they’ve never seen before. This is important because industrial processes are always changing, so the machine needs to learn from new information without forgetting what it already knows. The researchers propose a new way of doing this called incremental zero-shot fault diagnosis (IZSFD). They also create a special framework called BDMAFF that helps the machine learn and remember new information. This framework is tested using real-world data and shows that it’s effective in diagnosing unseen problems. |
Keywords
* Artificial intelligence * Zero shot
