Imagine three Russian dolls where Artificial Intelligence is the biggest, Machine Learning is inside it, and Deep Learning is the smallest inside both. This shows that every type of Deep Learning is also a form of Machine Learning and AI.

The Nesting Doll Hierarchy

Deep Learning is very 'hungry' for information and needs millions of examples to work well. While simple Machine Learning can learn from a small spreadsheet, Deep Learning needs massive amounts of data.

Large Scale Big Data

Deep Learning requires very fast and powerful computer chips, often called GPUs, to do millions of math problems at once. Traditional Machine Learning can usually run on a normal laptop without special equipment.

Specialized Computer Power

In Deep Learning, the computer automatically figures out what features are important, like the wheels on a car. In basic Machine Learning, a human usually has to tell the computer which features to look for first.

Automatic Feature Finding

The Power of Scale