A Response To @trumpman On How Neurons Are Converted To Zeros And Ones('0s' & '1s')

The basic unique feature which makes human to acts fast with accuracy and higher rate of pattern recognition, such as recognizing familiarized faces is the presence of biological neurons in human brain, therefore for computer to acts and carry out activities, fast and accurate like human brain, then it must be modeled according to human neurons, this is what led to the artificial neural network: a relatively new phenomenon in artificial intelligence field, which is an engine that drives the major progress in many fields such as medicine and entertainment.

credit

It is arguably undisputed that scientist are working days and nights, in order to completely model the neurons in human brain, starting from Artificial Neural Network to Convolutional Neural Network and Deep learning. In order to have a clear understanding of how Artificial Neural Network has helped in achieving major technological development in Artificial Intelligence, it then behoove me to discuss Artificial Neural Network and the science of how neurons are converted into zeros & ones as requested by @trumpman in his post here.

What is Artificial Neural Network?
Artificial Neural Networks are inspired by biological human neurons designed on a computing system to perform certain tasks.

How Does Artificial Neural Network Works?
Artificial Neural Networks was modeled to mimic human brain neurons by making the right connections, this can be achieved using silicon and wires as living neurons and dendrites in biological human neurons.
In reality, adult human brain consist of almost 100 billion nerve cells called ‘neurons’, which are connected parallel to another thousands of nerve cells by the so called ‘Axons’.
Stimuli/Signal from environment or inputs (inform of electric impulse) from sensory organs are accepted by dendrites. The inputs then travel down through the neural network for a neuron to send the message to another neuron within the connectionist.

credit

However, Artificial Neural Networks composed of multiple nodes, which mimic biological neurons presence in human brain. Just like the biological neurons, the neurons are connected and interact with one another through the interconnection link. In Artificial Neural Network, the node takes an input in form of data and perform simple activities on the data, the result is then pass over to other neurons. The output of each node is called ‘activation’ and each link is attributed with an associated weight.

Structure of Artificial Neural Network
Artificial neural Network consist of basic three layers, namely : Input layer, Hidden layer and Output layer as shown in the diagramm below

credit

In one of my posts here, i discussed that everything on a computer screen, including the text you are reading now is represented as an image: which consists of an array of values called 'pixels', each pixels consist of three bytes with three colors channels(RED, BLUE AND GREEN) and for the gray channel which consist of (BLACK AND WHITE) colors, Black denotes '0' and White denotes '1'. For instance, let us consider an image '3' with 28 by 28 pixels below

credit

The image '3' above has 28 by 28 pixels which serves as an input to the neural network with (28 * 28) pixels = 784 pixels i.e 784 of zeros and ones, since each pixel ranges from zero to one, thereby an image '3' here will serves as an input to be trained by neural network and it will consist of series of zeros and ones in 784 times.
To buttress my idea: without deepening much into theory of how Artificial Neural Network works, since there are many materials out there, I thereby recommend this video.

credit

I hope my little contribution here will enhance our basic understanding of how Artificial Neural Network works.

References
1
2
3
4
5
6
7

Thanks for reading through, your thoughts are very important!

In my next post, i will explore how Artificial Neural Network has helped in achieving major technological development in Artificial Intelligence.