How to make a neural network in Deep Learning?

Using Jupyter Notebook

This is very important project in which we’ll make a neural network. Neural network is very important thing in deep learning and machine learning mainly. The basic thing is that firstly, a neural network is made then it’s trained for a specific task. Here, we’re just making a neural network. So let’s start.

Firstly, we’ll import libraries.

importing libraries

Now we’ll create an artificial neural network called ANN. It is similar to biological neural network having input layer, cell biology(hidden layer) and output layer. Input layers receive input signals, hidden layers process inputs signals and finally output layer transfer it further to next level.

So, here’s a code to make an artificial neural network.

neural network

Here, relu, sigmoid, tanh are activation functions which are used to check the activation of neurons as well as for non-linearity. Here is optimizer ‘sgd’ that is standard gradient descent. It’s used for optimization means reduce error. ‘mse’ is mean square root value that shows loss/error.

Here is output.

output

So readers! here is a simplest way to make a neural network. If you really like my work, don’t forget to clap.

Thank You.