Basics of Hashing

AI Summary

An AI-generated summary of this video lecture.

This lecture provides an introduction to hashing, a data structure technique designed to optimize search operations. The instructor begins by contextualizing hashing within the broader scope of data structures, noting that while storage is the primary goal, search is the most frequent operation. He reviews the search complexities of various structures, such as O(n) for unsorted arrays and O(logn) for sorted arrays and AVL trees. The core of the lecture defines hashing as a method to achieve constant time O(1) search by mapping keys directly to memory addresses. It concludes by defining hash functions and introducing the concept of collisions, where distinct keys map to the same location, necessitating collision resolution strategies. The lecture sets the foundation for understanding how hashing improves efficiency over traditional linear or tree-based searches.

Chapters

1. 0:00 – 2:00 00:00-02:00

   The instructor starts by emphasizing that the most common operation in data structures is search, not insertion or deletion. He presents a slide listing search complexities for different structures: Unsorted array O(n), sorted array O(logn), linked list O(n), BT O(n), BST O(n), and AVL O(logn). To illustrate why these take time, he draws red diagrams next to the list. He sketches rectangular blocks for arrays and linked nodes for linked lists, showing arrows indicating a sequential traversal to find an element. He specifically crosses out 'Unsorted array' and 'BT' and 'BST' while keeping 'sorted array' and 'AVL' to contrast complexities. He underlines that search time depends on the number of elements and the structure type.

2. 2:00 – 5:00 02:00-05:00

   The topic shifts to hashing, defined on the slide as a technique where search time is independent of the number of items. The instructor explains the basic idea: using the key itself to find the memory address. He gives examples like phone numbers, roll numbers, or Aadhar cards. A diagram shows a list of keys on the left mapping to a table of buckets in the middle. He explains that a hash function converts a key into a smaller practical number to use as an index. He highlights the text 'access the element in O(1) time' with a red circle. He also points out 'overflow entries' in the diagram, showing how extra data is handled when a bucket is full. He draws a red arrow from a key to a bucket index, illustrating the direct mapping process. He mentions that the hash function must be modified so a great deal of space is not wasted.

3. 5:00 – 8:50 05:00-08:50

   The instructor formally defines the hash function as a conversion from a set of keys K to a set of memory locations L, written as H: K -> L. He explains that it maps a big number or string to a small integer. He then introduces 'Collision', explaining that two different keys K1 and K2 might yield the same hash address. He writes H(K1) and H(K2) on the board with arrows pointing to the same location L. He states that the technique to resolve this is called collision resolution. He draws a diagram showing n < m (keys less than locations) but notes collisions can still happen. He underlines the definition of collision on the slide. He explains that even if the number of keys is less than the number of locations, collisions are possible.

The lecture effectively bridges the gap between traditional data structures and hashing. It starts by establishing the problem: search operations in structures like arrays and trees are not constant time. It then proposes hashing as the solution, offering O(1) access. The instructor details the mechanism of hash functions and the inevitable issue of collisions, setting the stage for learning how to handle them. This progression moves from problem identification to solution definition and finally to the specific challenges of the solution, providing a comprehensive overview of the hashing concept.