Double Hashing

AI Summary

An AI-generated summary of this video lecture.

The video lecture provides a comprehensive introduction to Double Hashing, a collision resolution technique used in hash tables with open-addressing. The instructor, Sanchit Jain Sir, explains that double hashing uses a secondary hash of the key as an offset when a collision occurs. He contrasts this with linear and quadratic probing, noting that in double hashing, the interval depends on the data, which minimizes repeated collisions and clustering. The instructor writes the probe sequence formula $h_i(k) = (h_1(k) + i \cdot h_2(k)) \pmod m$ on the screen. He defines $h_1(k)$ as the primary hash function and $h_2(k)$ as the secondary hash function. He specifies that $h_1$ maps to $[0, |T|-1]$ and $h_2$ maps to $[1, |T|-1]$. The variable $i$ represents the probe number, starting from 0. The lecture emphasizes that $h_1$ and $h_2$ should be independent and uniform to approximate a random distribution. The slide text further explains that pair-wise independent hash functions yield a probability of $(h_1/|T|)^2$ that any pair of keys will follow the same bucket sequence.

Chapters

1. 0:00 – 2:00 00:00-02:00

   The instructor introduces Double Hashing, writing the formula $h_i(k) = (h_1(k) + i \cdot h_2(k)) \pmod m$ on the screen. He explains that unlike linear probing, the step size is determined by a second hash function $h_2(k)$. He writes $i=0, 1, 2$ to demonstrate the iteration. The slide text defines the technique and mentions minimizing repeated collisions. He details the ranges for $h_1$ and $h_2$ as $[0, |T|-1]$ and $[1, |T|-1]$ respectively. He also mentions that $h_1$ and $h_2$ are selected from a set of universal hash functions.

2. 2:00 – 2:10 02:00-02:10

   The instructor concludes the explanation of the formula's components. He highlights the independence of the hash functions to ensure a random distribution. He mentions that pair-wise independent hash functions yield a probability of $(h_1/|T|)^2$ that any pair of keys will follow the same bucket sequence, ensuring minimal clustering. He reiterates that this method approximates a random distribution effectively.

The lesson progresses from a high-level definition of double hashing to a detailed breakdown of its probe sequence formula. By writing the equation $h_i(k) = (h_1(k) + i \cdot h_2(k)) \pmod m$, the instructor clarifies how the secondary hash function determines the step size, effectively solving the primary clustering issues found in linear probing. The constraints on the ranges of $h_1$ and $h_2$ are crucial for ensuring that the entire table is probed.