In general, as the sample size n increases, do we expect the test prediction accuracy of QDA (Quadratic Discriminant Analysis) relative to LDA (Linear Discriminant Analysis) to improve, decline, or be unchanged? Why?

O Improve, because QDA is more flexible and can capture complex relationships.
O Decline, because QDA tends to overfit with larger sample sizes.
O Be unchanged, because QDA and LDA have the same predictive accuracy regardless of sample size.
O Improve, because LDA relies on strong distributional assumptions that become more accurate with larger samples.