AWS Certified AI Practitioner AIF-C01 Practice Question

When a model shows similarly low performance on both the training and validation sets, it is not capturing the underlying patterns in the data. This is a hallmark of high bias, which leads to underfitting. High variance and overfitting would show high training accuracy but much lower validation accuracy. Data leakage usually results in unexpectedly high validation accuracy. While label imbalance can reduce overall accuracy, it would not necessarily keep training and validation accuracy equally low if the model were sufficiently complex.