In the knearest neighbor classifcation framework. K designates the number of nearest neighbors. As K increases, then typically C] 1.the variance increases and the bias decreases. C] 2. the prediction power becomes higher as we take into account more neighboring points. C] 3. the training mean squared error increases and the test mean squared error rate decreases. C] 4. the bias increases and the variance decreases. C] 5. the test error rate decreases and the training error rate increases