Imagine you had a learning problem with an instance space of points on the plane and a target function that you knew took the form of a line on the plane where all points on one side of the line are positive and all those on the other are negative. If you were constrained to only use decision tree or nearest-neighbor learning, which would you use? Why?

The answer is nearest-neighbor learning.

because nearest neighbor learning is classification algorithm.

It is used to identify the sample points that are separated into different classes and to predict that the new sample point belongs to which class.

it classify the new sample point based on the distance.

for example if there are two sample points say square and circle and we assume some center point initially for square and circle and all the other points are added to the either square or circle cluster based on the distance between sample point and center point.

while the goal of decision tree is to predict the value of the target variable by learning some rules that are inferred from the features.

In decision tree training data set is given and we need to predict output of the target variable.

Explanation: