Neural Networks

A biological neural network is a collection of connected neurons that offer some function as a nervous system or brain. But the term “neural network” usually refers to a software logic array that attempts to mimic some part of the design or function of a biological network. In the biological form, there are neuron cells that have some ability to process signal form and the cells are connected by gaps called synapses, which involve bio-chemical transfers to get a signal from one cell to another.

Artificial neural networks are inspired by the biological form and use clusters of mathematical formulas to represent the neurons and connect them in array-like structures. Raw input data is fed into one part of the array, which is designed to process the data and create some form of output on another part of the array. The array is then “trained” (using error feedback) by tweaking the processing formulas to create the desired output data. As the array is continually trained, the output gets closer and closer to the goal until it reaches a performance standard that is acceptable, according to the function of the network. The array can now process un-analyzed data that fits into the training profile and realiably creates accurate output.

Neuron firing – biological neurons “fire” or transmit information when the signal they receive reaches a certain threshold. Artificial neurons work the same way, but using formulas to calculate when they should fire and transmit information along the network.

Feedback – one common form of neural network, known as a “back propagation” network, uses feedback from output errors to adjust the firing thresholds and train the network until it approaches a desired level of accuracy. Once this training operation has been completed, the network can be used for actual production tasks. This, however, is not a live feedback training process in most networks, with discrete training modes and operational modes that do not operate simultaneously.



This report is an introduction to Artificial Neural Networks. The various types of neural networks are explained and demonstrated, applications of neural networks like ANNs in medicine are described, and a detailed historical background is provided. The connection between the artificial and the real thing is also investigated and explained. Finally, the mathematical models involved are presented and demonstrated.

Leave a Reply

You must be logged in to post a comment.