There is a relatively new approach to addressing large, complex problems - 'network science'.
Dr John Finnigan at CSIRO Atmospheric Research says network science is gaining recognition as a more efficient and productive means of solving complicated problems than the traditional 'examine-one-aspect-at-a-time' approach.
"Network science is applied in areas as diverse as: social organisation, traffic and pedestrian flows, food web behaviour in natural and artificial environments, genetic control and cancer spread in cells, disease propagation and data security and transfer in hostile environments like war zones," Dr Finnigan says.
"Network give us radical insights into many areas. For example, the power blackouts that have crippled parts of Australia, the North-Western USA, and Italy in recent years are not caused by exceptional one-off events as was first thought," Dr Finnigan says.
"In fact, network science shows that such events can be expected because of the pattern of connections in the power grid," Dr Finnigan explains. "We can now predict the patterns that allow failures to cascade through the power grid instead of being confined to a small area and rapidly corrected. Knowing the patterns helps us to predict and manage trouble spots.
"More surprisingly, we know that the same rules govern the spread of ideas through a population or the uptake of a new technology like DVDs or iPODs, the success of businesses, and the failure of communication in organisations."
Dr Finnigan says network science can also be a useful means of investigating other phenomena such as in identifying the key people to arrest or discredit in order to disrupt criminal networks or terrorism cells.
Known as Pattern to Knowledge (P2K), the technology uses artificial intelligence to leverage deep...
Stanford scientists have developed an algorithm, called Phrank, that automates the most...
By analysing the brain activity of tetraplegia patients, a new deep-learning algorithm can help...