Neuron Rank: The Browser of the Brain
Both the brain and the internet can be represented as a graph. The nodes that compose the brain are neurons while the nodes that compose the internet are pages. The edges that compose the brain are synaptic connections while the edges that compose the internet are hyperlinks. Even more important, both the internet and the brain are more than the sum of their parts. In others words,the edges tell a story that nodes ignore.
Both the internet and the brain, must be able to efficiently find what it is needed in gigantic amounts of data, otherwise they become useless. The way that internet handles this problems is through the use of browsers which implement methods to achieve this task. One of theses methods is to rank pages through the Hubs and Authorities algorithm.
According to an article published by Science Direct titled “Ranking neurons for mining structure-activity relations in biological neural networks: Neuron Rank,” brain activity may partially modeled by the Hubs and Authorities algorithm, a process that is referred in the article as “neuron rank.”
In this model each neuron is idealized as a simple independent entity. All it can do is to add the signals it obtains from the incoming edges (dendrites) ,and output a signal through the axon.
In order to compare a neural network to a hub and authorities graphs, we need to assume that the brain has specific regions that analyze stimuli from the outside world ,and specific regions that decide what action to take as a response to those stimuli. If we arrange the neurons that receive stimuli on one side, and the neurons that carry out a response on the other we obtain the following graph:
(Notice that the layer in the middle is just to simplify the view of the model and does not do any computations).
In order to be more concrete, we can think of the possible actions that an animal (or a plant) can take (transpiration, assimilation, etc.) to the possible related stimuli (light, temperature, etc.) Every time a living organism interacts with its environment a set of stimuli will be linked to a set of decision. This can be compared to the relationship between hubs and authorities.
As the cycle of perception-decision progresses a state of equilibrium will be reached, after which a clear rank of the decision will be establish. As a result, the living organism will have a “rank” of possible responses given certain stimuli. This perception-decision cycle is very similar to the Hubs and Authorities endorsing process. Both process favor more connected nodes, and go through a never-ending endorsing cycle that reaches equilibrium. At the end, the top option is the more likely option to be taken. But unlike the internet browser, you cannot put an advertisement next to the top option of the mind (yet).
References
Neuron rank research papers
https://lirias.kuleuven.be/bitstream/123456789/124709/1/neurocomputing06.pdf
http://wwwconference.org/www2005/cdrom/docs/p936.pdf
Neural networks description (from which the first image was taken)
Intro to neural network (from which second image was taken).
http://lowercolumbia.edu/students/academics/facultyPages/rhode-cary/intro-neural-net.htm
-Near Rivet