#bias-in-machine-learning

[ follow ]

The Specifics Of Data Affect Augmentation-Induced Bias | HackerNoon

Excessive data augmentation can induce significant bias in model performance, differentiating among various data classes.

Countering Mainstream Bias via End-to-End Adaptive Local Learning: Preliminaries | HackerNoon

The study addresses mainstream bias in recommender systems by aiming to improve utility for niche users while preserving utility for mainstream users.

(Practical) AI ethics

Bias in machine learning is intrinsic and must be managed by determining acceptable biases.
[ Load more ]