Summary of Fair Class-incremental Learning Using Sample Weighting, by Jaeyoung Park et al.
Fair Class-Incremental Learning using Sample Weighting
by Jaeyoung Park, Minsu Kim, Steven Euijong Whang
First submitted to arxiv on: 2 Oct 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 In this paper, researchers investigate the problem of unfair catastrophic forgetting in class-incremental learning for Trustworthy AI. While accuracy has been a primary focus, fairness is relatively understudied. The authors theoretically analyze that forgetting occurs when the average gradient vector of the current task data is in an “opposite direction” compared to the average gradient vector of a sensitive group, leading to unfairness. To address this issue, they propose a fair class-incremental learning framework that adjusts training weights to reduce forgetting and achieve fairness for various group fairness measures. They formulate optimization problems using linear programming and propose an efficient Fairness-aware Sample Weighting (FSW) algorithm. Experimental results show that FSW achieves better accuracy-fairness tradeoff results compared to state-of-the-art approaches on real datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper explores how machines learn in a way that is fair for everyone. Right now, machines are not always fair when they learn from new data. The authors figure out why this happens and propose a new way of learning that makes it more fair. They test their method on real data and show that it performs better than other methods. |
Keywords
* Artificial intelligence * Optimization
