the neural network or artificial neural network (ANN) defined by DR. Robert Hecht-Nielsen the inventor of the first neurocomputer as:
“A computing system made up of a number of simples. Highly interconnected processing elements. Which process information by their dynamic state response to external inputs”
What is the neural network?
The neural network is a computing system. Inspired by the biological neurons in the animal brain. So, these networks learn new things by considering examples. Unlike regular coding that requires task-specific programming. Neural networks can learn and identify similar items in the future. So, let us explore what is a neural network in a simple way.
Neural networks example:
As we mentioned, neural networks can learn from examples and apply later. So, neural network recognizes things you already taught how to recognize it but with more flexibility. So that, If you want it to select cat picture from different animals pictures. Then, all you have to do is providing some cat picture examples and let it do the magic.
The neural network will understand the example pictures and analyze it to elements such as fur, ears, eyes etc. When you apply this to different animals it will select the cat even with new cats pictures.
How does neural network work?
So, They organized ANNS layers. Then, they made each layer of several “nodes” that contain activation function. So, you represent models to the network through input layer. And, they did organize the ANNS in layers. And, they made Each layer of several “nodes” that contain activation function. You represent models to the network through input
So then, The input layer communicates to hidden layers, then link to the output layer. Hidden layers do The actual process via the system of weighted connections. Usually, ANNs have a form of ‘learning rule’. That modifies the weight of each connection according to the input model that it present with. So, ANNs learn by experience(examples), the same way Animals learn.
What programs use neural networks?
Neural networks are universal technology, you can use it in infinite situations. But the ideal situation for it to work. Is that when the model you are working on has a high tolerance to error. Of course, I don`t advice to use neural networks in a simple task like a calendar for example!
Also, There is the G.A.N. technology As known as pix2pix. A tool that you give some outlines. Of a cat or human face or whatever you are willing to draw as an input. And take back a fully drawn picture as an output. You don’t give information about that element to draw or anything. But Neural network finishes it all for you.
- Selecting similarities within a set of patterns.
- If the number or the volume of data is so great.
- To understand the relationship between variables.
- The relationships are difficult to describe adequately with conventional approaches.
Also, Using the technology for robots machine learning. Will ensure having a new generation of smart robots. So, it changes the future of A.I.
- The neural network is a computing system. Inspired by the biological neurons in the animal brain. These networks learn new things by considering examples.
- Also, The Neural network is able to understand by itself and recognize what you don’t give her information about.
- And the neural network using 3 type of layers. Input, hidden, And output layers.
- It is not wise to use it in a simple program or task. But we should use it in huge projects or big programs that we face enough complication in it.
- Using this technology with robots and drones. Will create a new era of robotics.