AI Engineering Degree Practice Exam 2025 - Free AI Engineering Practice Questions and Study Guide

Increasing the number of neighbors (k) in a K-Nearest Neighbors (KNN) model tends to reduce variance in the model's predictions. This is because a larger value of k means that the model is considering a broader group of data points when making predictions, which leads to a more stable output by averaging over more instances.

As k increases, the model becomes less sensitive to the noise and fluctuations in the training data, thus smoothing out the predictions and reducing the effect of any outliers. This helps in combating the overfitting problem, where a model becomes too complex and captures noise rather than the underlying pattern in the data.

While increasing k does not eliminate errors entirely and it might not improve bias, which is the tendency of a model to consistently miss the mark, it generally leads to a more generalizable model on unseen data. Therefore, the correct choice indicates that increasing k is effective in lowering the model's variance, promoting more consistent predictions across diverse datasets.