Given a large ordered array that may contain many elements that are multiply occurring (it may be assumed that on average when an element is multiply occurring it will occur many times). Devise an adaptive duplicate deletion algorithm that is more efficient than the above algorithm for this particular type of data.