How Sensitive Data Affects Fairness and Accuracy in Medical AI Models | HackerNoon
Briefly

from Hackernoon1 month ago
The impact of using sensitive attributes on model fairness and performance is significant, as demonstrated by the comparative analysis of four models in chest radiography classification.
Hackernoonhttps://hackernoon.com/how-sensitive-data-affects-fairness-and-accuracy-in-medical-ai-models
In our study, M2 incorporates race features, suggesting that explicit use of sensitive attributes can enhance model performance, while M3 aims to exploit inherent racial encodings.
Hackernoonhttps://hackernoon.com/how-sensitive-data-affects-fairness-and-accuracy-in-medical-ai-models
Read at Hackernoon
#fairness #machine-learning #chest-radiography #model-comparison #sensitive-attributes
[
|
]