Unsupervised Learning is the training of machine without any labeled data unlike (supervised learning). Basically, in unsupervised learning, there is no classified data and the algorithm needs to work without any guidance or training. The model has to group unsorted information using similarities, and differences.
Let me explain this better with a very cliche example:
Suppose you have a fruit basket which consists of 4 different fruits - Apple, cherry, banana, and grapes. As mentioned above, unsupervised machine learning does not have predefined data or labeled data. So how will the machine group these fruits?
It will start categorizing them based on differences and similarities.
Suppose it starts categorizing them with respect to colors-
- red - apple and cherry
- green - banana and grapes
It will next start categorizing them based on size
- red and small - cherry
- red and big - apple
- green and big - banana
- green and small - grapes
Yyayyy congratulations! categorizing done without any training!