Understanding Neural Networks
In this tutorial, we’ll understand the basics of neural networks, like what is a neuron, what are neural networks, the architecture of neural networks, the classification of neural networks, etc.
Neural networks in Deep learning works similar to the human brain. Most of the Deep learning models are trained using large set of data and neural network architectures.
What is Neuron?
-
Neurons are the foundation of the neural networks.
-
As, we already mentioned these neurons are similar to the biological neurons inside human brain.
-
In technical terms, neuron is a mathematical function that receives one or more inputs and performs a computation on these inputs, and produces an output.
What is Neural Network?
-
Neural networks in deep learning are a layered model that is inspired by the structure and function of the human brain.
-
These neural networks consist of interconnected neurons, that are used to process information and learn to recognize patterns in data.
-
These neural networks are trained on large datasets to recognize patterns.
-
These neural networks consist of an input layer, some hidden layers, and output layer.
Architecture of a Neural Network
-
A neural network is made up of Layers of connected nodes, or neurons.
-
One or more inputs are taken by each neuron, and it then processes them to form some outputs.
-
These outputs can be used as the input to another neuron, allowing information to flow through the network.
How neural networks work?
-
In Tensorflow, we can create and train neural networks with the help of an high level API known as keras.
-
To create a neural network in tensorflow first we have to define its architecture, number of neurons in each layer and activation function.
-
Then, we have to specify the optimizer used for compilation, the loss function and the metrics we want to track during the training of the neural network.
-
At last, we can feed data and test its performance on the given data.
In upcoming tutorials, we’ll learn about the classification of these neural networks.
