Main menu



Artificial Neural Networks (ANNs)

An artificial neural network (ANN) is a type of computing system that has patterns created by mimicking the actions of neurons in the human brain.

How do artificial neural networks work?

An artificial neural network (ANN) is best viewed as a weighted directed graph. This is usually structured in several layers. In this layer there are numerous nodes that mimic the biological neurons in the human brain, which are interconnected and contain activation functions. The first layer receives raw (raw) input signals from the outside world. It is similar to how humans use the optic nerve to process visual information. Subsequent successive layers each receive the output from the preceding layer, similar to how neurons farther away from the optic nerve receive signals from the nearest optic nerve. The output of each node is called its activation or node value. The last layer derives the output of the system. ANNs are actually mathematical models that can learn. Thus, ANNs have been used to augment existing data analysis techniques. ANNs are one of the reasons behind the significant advances in artificial intelligence (AI), machine learning (ML) and deep learning.

Perceptron artificial neural network

A Perceptron is the simplest type of artificial neural network. This type of network is usually used to make binary predictions. Perceptrons only work if the data can be separated linearly.
What is Artificial Neural Network in 5 min?

multilayer artificial neural network

A fully connected multilayer neural network is also called a Multilayer Perceptron (MLP). This type of artificial neural network consists of one or more layers of artificial neurons or nodes (such as a Convolutional Neural Network or Recurrent Neural Network). Multilayer ANNs are used to solve complex classification or regression tasks. The most common model is the three-layer fully coupled error backpropagation model. The first layer consists of input neurons, which send data to the second layer, which in turn sends the output neurons to the third layer.

There are two other artificial neural network (ANN) topologies. FeedForward and Feedback.

FeedForward artificial neural network

In this ANN, the flow of information has no direction. Information only travels in one direction, forward and does not create a feedback loop. The information will first pass through the input node, then through the hidden node (if any), and finally through the output node.
FeedBack artificial neural network

In this case, an intrinsic feedback connection exists between the multiple neurons in the network. A feedback loop is allowed here.