What is an Artificial Neural Network in Data Science?

Data Science, AI and ML
,
#1

What is an Artificial Neural Network in Data Science?

#2

An artificial neural network is the computing system made up of a number of simple, highly interconnected processing elements, which process information by their dynamic state response to external inputs.

#3

Neural networks are programs that try to mimic the way brains work. They simulate what it is like to have millions of neurons (the cells that make up the brain) as they learn to recognize patterns. Just like a biological brain, you need to teach a neural network for it to learn a new skill. For example, you might show a neural network hundreds of photos of kittens if you want it to recognize kitten photographs.

Neural networks differ from biological brains in some important ways:

  1. Most neural networks can only learn one thing. So a neural network you train to recognize kitten photos would never be able to recognize photos of cars.
  2. Neural networks don’t “think” in the way you and I do. They only recognize common patterns, but cannot do any form of reasoning.

The limitations are a result of our lack of understanding about how brains actually work, and how complex they are for what we do understand. Most neural networks we can build today are less intelligent than a two year old child.