Summary of Beyond Rmse and Mae: Introducing Eauc to Unmask Hidden Bias and Unfairness in Dyadic Regression Models, by Jorge Paz-ruza et al.
Beyond RMSE and MAE: Introducing EAUC to unmask hidden bias and unfairness in dyadic regression models
by Jorge Paz-Ruza, Amparo Alonso-Betanzos, Bertha Guijarro-Berdiñas, Brais Cancela, Carlos Eiras-Franco
First submitted to arxiv on: 19 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
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 The paper proves that non-uniform observed value distributions of individual entities lead to severe biases in state-of-the-art dyadic regression models, skewing predictions towards the average of observed past values for the entity. This bias, named eccentricity bias, results in worse-than-random predictive power in crucial cases. The study introduces Eccentricity-Area Under the Curve (EAUC) as a novel metric to quantify this bias and evaluates its effectiveness in various domains. By introducing EAUC, the work contributes to developing fair models that prevent unfairness in critical real-world applications of dyadic regression systems. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper shows that some machine learning models are biased when they try to predict things like how well a patient will respond to a certain medicine or what kind of product someone will like. This bias can make the predictions worse than if the model just guessed randomly! The researchers found that this problem is caused by the way data is distributed for each individual thing being predicted, like patients or products. They also came up with a new way to measure how well these models are doing, called EAUC (Eccentricity-Area Under the Curve). This can help us make sure our models are fair and don’t discriminate against certain groups. |
Keywords
* Artificial intelligence * Machine learning * Regression
