What does come immediately in your mind when you hear of the word Neural Network?
something related to or kind of NEURONS, Right? You are somewhere near it, but it is not at all related to humans.
Just after the human brain, neural network are a set of algorithms that are modelled loosely , and are designed to recognize patterns. They interpret sensory data through a kind of machine perception.
Labelling or clustering raw inputs.
Neural Network forms the base of deep learning, which comes under machine learning where the algorithms are inspired by the structure of the human brain.
Neural Networks take in data, train themselves to recognize the patterns in this data and then predict the outputs for the new sets of similar data.
Lets understand how this is done :
Neural Networks are made of the nets of neurons, this Neurons are the core processing units of the network, first we have the input layer which receives the input, the output layer predicts the final output, in between exists the hidden layer which perform most of the computations required by our network.
Each neuron of first layer is attached to the second layer through channels each channel is assigned a numerical value known as weight. The inputs implied to the corresponding weights and their sum is send as input to the neurons in the hidden layer. Each of this Neurons is associated within the numerical value called the bias. Which is than added to the input sums. This value is than passed to a threshold function, called the activation function. The result of the activation function determines that the particular neuron will get activated or not, an activated neuron transmits the data to the neuron to the next layer over the channels, in this manner the data is propagated through the network, this is called forward propagation, in the output layer the neuron with the highest number fires and determines the output, the values are basically a probability, and that is the output predicted by the neural network. The predicted output is compared against the actual output to analyse the mistake in the prediction.
This network is than transferred back through a network, this network is known as backpropagation.
Based on this propagation the weights are adjusted, this operation of forward propagation and back propagation are iteratively performed with the multiple inputs, this activity continues until the weights are assigned such that the network can predicts the answer correctly. In most of the cases this brings our training process to an end.
You might wonder how much time does this training process takes, honestly Neural Network may take an hour or months to train. But time is a reasonable trade-off when compared to its scope.
Let us look at some of the prime applications of the neural network :
1)Facial recognition:- now a days the smartphones in your hands can forecast your age , this is where Neural Network come at work, firstly they differentiate the face with the background, then make a Neural Network in face and by recognizing the facial features try to predict the age.
2)Forecasting:-neural network are trained to detect the problems and detect the possibilities of rainfall , water rise, crisis with high accuracy.
3)Music composition:- neural network can even learn music and tunes and train itself to compose a fresh tune.
Hope you have now understood that what is NEURAL NETWORK.