1.) Suppose you are trying to predict the height of an animal in afield. 9 out of 10 animals you see are mice. 1 out of 10, however,is a deer, whose height is 20 times that of a mouse. Your estimatoris the average height of all animals (let m be the number ofanimals that you have observed). Does the data have a large bias,or a large variance, or both? Clearly explain your answer. Nocomputation is needed.

2.) Suppose now you take the mouse height as a predictor for theaverage height (this makes sense since mice are much more commonthan deer). Would the bias go up or down, compared to your answerto the previous question? Did the variance go up or down? Clearlyexplain your answer.

Answer rating: 100% (QA)

1 Bias and Variance in the Average Height Estimator In this scenario the estimator for predicting the height of an animal is the average height of all ... View the full answer