Neural Network

Do you remember when you attended your first math class? You were unaware of additions & subtraction before it was taught to you. But today you can do it at your fingertips. This was possible only due to a lot of practice! All the gratefulness to our highly complex brains with billions of interconnected nodes (called neurons) that we can keep learning stuff.

Well, the concept of Neural Network is based on the same principle as our brain, even though it is impossible to replicate our brain, we can at least create a simple miniature model of it which can learn to perform the single task efficiently and then use that replicated brain to perform daily tasks.

Just like our brain contains neurons and synapses connecting them, Neural Networks also contain neurons, and the connection between these is called weights. Just like our sensory system sends our brain signal, Neural Network also sends the signal back using something known as backpropagation. Just as we improve our mistakes by comparing our actions with the right actions, Neural Networks also compare their results with actual results using something known as the cost function. Just like our brains tell us to do a specific action in a particular situation by sending signals to our muscles, Neural Networks also send signals using the process known as feedforward propagation.

This is just a basic example of the structure of one type of neural network. Yes, there are other types of Neural Networks as well, and we are going to discuss them in this course.

We will first start with a brief introduction to the concept of Neural Networks and the mathematics behind them and then continue looking at the different applications of Neural Networks using MATLAB and its Neural Network Toolbox.