Bloom filters use a bit vector and multiple hash functions to test set membership quickly and memory-efficiently.
They can definitively rule out non-members, but may return false positives for members.
The false positive rate is approximated by (1−e^(−kn/m))^k and can be tuned by choosing filter size and number of hashes.
The optimal number of hash functions k is (m/n) ln(2) for a given filter size m and expected number of elements n.
Insertion and membership tests run in O(k) time, where k is the number of hash functions used.
Recommended hash functions include fast, non-cryptographic ones like Murmur, xxHash, and FNV.
Bloom filters are widely used in systems like Chromium, RedisBloom, SQLite, and RocksDB for efficient membership checks.
Get notified when new stories are published for "General AI News"