We know that hash function generates a unique hash value for a given data. This is similar to finger print. But there is a possibility that two data sets may generate same hash value. This is known as collision. 

How to find such collisions ?