Hello, guys it’s been a long time. In my previous posts we have seen some basics about machine learning and it’s application as part of Artificial Intelligence. We also know about the Natural Language Processing. In this post we are going to learn about neural network actually Artificial Neural Network(ANN) is more preferable.
Artificial Neural Network is a field of Artificial Intelligence and it is actually quite interesting one. It is an attempt to simulate the neural network structure like that of a human brain to do the computational work.
Brain is the most important part of the body and it is the one of the most complex organ in the world. Brain works as command center or call it a CPU of body where it receives input from the sensory organs and sends output to the muscles of body through nervous system. And how exactly brain process the data? brain is made of 100 billion or more than that nerve cells also called neurons. Neurons communicate with other neurons sending electrochemical signals via Axon. Generated signals are send back to the body muscles through nervous system. I'm not going to tell about the biological stuffs. But the concept of Neural Network or NN is inspired from the biological neural network. So, I think we need to know how the neurons works in term of biology.
In 1943, first time Neural Network came into picture after introduction to the simplified version of neurons like biological neurons and computational neuron circuit to perform computational task by McCulloch and Pitts. This event caught many computer scientists' eye to do research and development in this field.
So, what is Artificial Neural Network?
In computer science an artificial neural network is an interconnected group of nodes called as neurons, which is like a vast network of neurons in a brain for computation process in parallel manner. below picture shows the simple structure of neuron which could have 1 or more input but number of output will never change it's always 1. Xi is input, where i = N(natural number) and y is output.
Don't be confuse about neuron. It is nothing but a function or operation for performing a particular task. It's like a function in programming language.
Neural Network is often multilayered and in each layer several neurons are ready for processing.
Above diagram illustrate the multilayer Neural Network where the first layer is called as 'Input Layer' in which initial input is give to the NN and last layer is called 'Output Layer' where final output returns from the NN. Layers between the i/p layer and o/p layer is called as hidden layers. Complexity of the Neural Network is measured by the number of hidden layers and the number of connections between the neurons. In diagram there are 4 layer neural netwok with 2 hidden layer and total 10 neurons and 4 initial input. Output of the first layer neurons is passed to the next layer neurons and so on. it is type of feed forward neural network.
Here, above figure shows the neural network of alphaGo program developed by Google DeepMind especially for GO game., where there are around 250 possible steps after one move which lead to the success. So in the figure there are around 250 output neuron and only one of them selected for further process.
And at last there is a program called Critterding that is an Artificial life simulation and it is based on neural network. I think it is a good example for Neural Network.
Artificial Neural Network is a growing field and I think this is what that make computation process more fast and efficient and so that Artificial Intelligence becomes more intelligent.
So that's all for this post and hope it is useful. If you liked it, share and subscribe our facebook page and Youtube channel for more update. You can follow us on Twitter and Instagram.
Artificial Neural Network is a field of Artificial Intelligence and it is actually quite interesting one. It is an attempt to simulate the neural network structure like that of a human brain to do the computational work.
Brain is the most important part of the body and it is the one of the most complex organ in the world. Brain works as command center or call it a CPU of body where it receives input from the sensory organs and sends output to the muscles of body through nervous system. And how exactly brain process the data? brain is made of 100 billion or more than that nerve cells also called neurons. Neurons communicate with other neurons sending electrochemical signals via Axon. Generated signals are send back to the body muscles through nervous system. I'm not going to tell about the biological stuffs. But the concept of Neural Network or NN is inspired from the biological neural network. So, I think we need to know how the neurons works in term of biology.
In 1943, first time Neural Network came into picture after introduction to the simplified version of neurons like biological neurons and computational neuron circuit to perform computational task by McCulloch and Pitts. This event caught many computer scientists' eye to do research and development in this field.
So, what is Artificial Neural Network?
In computer science an artificial neural network is an interconnected group of nodes called as neurons, which is like a vast network of neurons in a brain for computation process in parallel manner. below picture shows the simple structure of neuron which could have 1 or more input but number of output will never change it's always 1. Xi is input, where i = N(natural number) and y is output.
Don't be confuse about neuron. It is nothing but a function or operation for performing a particular task. It's like a function in programming language.
Neural Network is often multilayered and in each layer several neurons are ready for processing.
Above diagram illustrate the multilayer Neural Network where the first layer is called as 'Input Layer' in which initial input is give to the NN and last layer is called 'Output Layer' where final output returns from the NN. Layers between the i/p layer and o/p layer is called as hidden layers. Complexity of the Neural Network is measured by the number of hidden layers and the number of connections between the neurons. In diagram there are 4 layer neural netwok with 2 hidden layer and total 10 neurons and 4 initial input. Output of the first layer neurons is passed to the next layer neurons and so on. it is type of feed forward neural network.
Here, above figure shows the neural network of alphaGo program developed by Google DeepMind especially for GO game., where there are around 250 possible steps after one move which lead to the success. So in the figure there are around 250 output neuron and only one of them selected for further process.
And at last there is a program called Critterding that is an Artificial life simulation and it is based on neural network. I think it is a good example for Neural Network.
Artificial Neural Network is a growing field and I think this is what that make computation process more fast and efficient and so that Artificial Intelligence becomes more intelligent.
So that's all for this post and hope it is useful. If you liked it, share and subscribe our facebook page and Youtube channel for more update. You can follow us on Twitter and Instagram.
Comments
Post a Comment