Select all that hold true about Bloom Filters: Select one or more: A. Bloom Filter has a predictable false positive rate. (Incorrect choices are penalized) B. The total storage required for a Bloom Filter is fixed to be m bits, and this won't increase regardless of the number of values stored. C. Each search operation can be processed in the order of O(n x k) by a Bloom Filter with k hash functions that already added n values. O D. Each search or insert operation only requires O(k) operations by a Bloom Filter with k hash functions and the cost does not depend on either the number of bits used by the Bloom Filter or n, the number of values already added. E. While the size of the bit array used by the Bloom Filter is constant, it may need to be quite large to reduce the false positive rate. F The Bloom Filter supports the following operations: add/insert the element, search the element, and remove the element. G. While risking false positives, Bloom Filters have a strong space advantage over other data structures for representing sets, such as self- balancing binary search trees, tries, has tables, simple arrays or linked ists. OH. The Bloom Filter also stores the data items themselves that could be retrieved in O(k) time