In this problem we want to learn the concept class of origin$-$centered concentric circles.

More specifically, fix two circles $($both have centers at the origin$)$ in the plane. The learner

is given draws from a distribution where each point in between the two circles is labeled

positive, and every other point is labeled negative $($points on the boundary of either circle

are labeled positive$).$

$($a$)$ Give an algorithm for PAC learning this concept class. You may just describe it in

words or via pseudocode.

$($b$)$ Describe the bad events for your algorithm $($i$.$e$.,$ the events where your algorithm will

fail to output an accurate hypothesis$).$ Be as formal as you can.

$($c$)$ Now in terms of $lon$ and $$ analyze the sample complexity of your algorithm.