Neurone Theory

The Neurone theory of the Brain

Neurone illustration adapted from an original by Bruce Blaus

The brain is made up of individual cells that contain specialized features such as: the nucleus, dendrites , a cell body and an axon and Synapses These are called neurone units or cells.

The average human brain contains around 100 billion neurons – this is a HUGE number which we will return to when looking at neural nets later on.

The Neurone cells are connected by sites of contact through the synaptic terminals. Although the axon can conduct in both directions, here there is a preferred direction for transmission from cell to cell.

A barrier to transmission exists at the site of contact between two neurons through the synapses that may permit transmission. However if a contact is made between two cells, then that contact can be either excitatory or inhibitory, but will always be of the same type.

In simple terms; signals can pass from one cell to another cell l through the Axons. It will then be transmitted to the next cell thorough the synapses. However, it should be noted that neurone cells will either ‘fire’ – pass a signal or will inhibit a signal

As billions of cells fire – or not – when receiving information such as an image, sound, taste or touch – then we can see that a pattern will form in the brain as neurones fire on or off – the pattern of firing will be our internal representation of that information.

A simple example – Imagine an ice cream in a cone. When you ‘see’ the ice cream your neurones will fire in response to the image of the ice cream. Now the neurons in your brain will have a different firing pattern to mine – However, we both ‘see’ the same thing. Perhaps you do not like Ice cream, in which case your neurones will definitely fire differently to mine…

Now – think of two ice creams – One a chocolate ice and one a vanilla ice – you may think that all the same neurons will fire off as in the example above except that the ‘colour’ associated with each of the ices will be different – this may or may not be so – It may even be found that one set of neurons fire off at the Vanilla Ice and a totally different set will fire off at the Chocolate Ice. At this moment in time, no one knows – the only thing to remember here is; The brain and its workings and the patterns that form from the firing of the neurones , are not always quite as straightforward as you may think…

However, for our purposes, the exact manner or pattern of firing in the brain is not totally the issue. Neural networks, that seek to mimic the brain, they too exhibit a similar counter-intuitive firing pattern in their behavior. And this is exactly what we are looking for in machine learning – it is these things that make us ‘human’ and it is these ‘peculiarities’ we are seeking to emulate in Machine Learning.

Now jump to the next section to see Machine Learning in progress – In The Perceptron or, you can just Click Here.