selection bias

Selection bias occurs when the data used to train a model is not representative of the underlying population, leading to skewed results. Addressing selection bias is critical for ensuring model fairness, reliability, and generalizability in AI applications.

6 papers