Self-Organized Criticality

Dynamical systems that have a critical point as an attractor are called Self-Organised Critical.

Properties
Dynamical systems that are self-organized critical follow a power law distribution. Using this measure we can determine whether systems are self-organized critical or not.

Advantages
For neural networks that are SOC, there are certain advantages.
 * Optimal information transmission
 * Optimized imformation storage
 * Increased computational power
 * More stable

Applications
There has been research that suggests that neural networks that lack self-organized criticality in certain parts of the network may lead to ailments like epilepsy. Other research has shown that temporarily decreasing the SOC plays a a role in in sleeping. And that sleeping is critical in returning the system back to it's critical point.