Tabulation Hashing
In lecture, we’ve taken it as a given that we have good hash functions available. But how exactly do you go
about building these hash functions? One popular approach – which has the endorsement of CS legend
Don Knuth – is tabulation hashing, which breaks a key apart into multiple blocks and uses table-based
lookups to compute a hash code. Although tabulation hashing only gives 3-independence, which doesn’t
sound all that great, a deeper analysis shows that using this hashing strategy in certain contexts will per-
form much better than what might initially appear to be the case.
Do'stlaringiz bilan baham: |