"The first attempts to consider the behavior of so-called 'random neural nets' in a systematic way have led to a series of problems concerned with relations between the 'structure' and the 'function' of such nets. The 'structure' of a random net is not a clearly defined topological manifold such as could be used to describe a circuit with explicitly given connections. In a random neural net, one does not speak of 'this' neuron synapsing on 'that' one, but rather in terms of tendencies and probabilities associated with points or regions in the net." (Anatol Rapoport, "Cycle distributions in random nets", The Bulletin of Mathematical Biophysics 10(3), 1948)
"In a random network the peak of the distribution implies
that the vast majority of nodes have the same number of links and that nodes
deviating from the average are extremely rare. Therefore, a random network has
a characteristic scale in its node connectivity, embodied by the average node
and fixed by the peak of the degree distribution. In contrast, the absence of a
peak in a power-law degree distribution implies that in a real network there is
no such thing as a characteristic node. We see a continuous hierarchy of nodes,
spanning from rare hubs to the numerous tiny nodes. The largest hub is closely
fol - lowed by two or three somewhat smaller hubs, followed by dozens that are
even smaller, and so on, eventually arriving at the numerous small nodes."
"Networks are not en route from a random to an ordered state. Neither are they at the edge of randomness and chaos. Rather, the scale-free topology is evidence of organizing principles acting at each stage of the network formation process."
"[…] real networks not only are connected but are well beyond
the threshold of one. Random network theory tells us that as the average number
of links per node increases beyond the critical one, the number of nodes left
out of the giant cluster decreases exponentially. That is, the more links we
add, the harder it is to find a node that remains isolated. Nature does not
take risks by staying close to the threshold. It well surpasses it."
"Regular graphs are unique in that each node has exactly the same number of links. […] Such regularity is clearly absent from random graphs. The premise of the random network model is deeply egalitarian: We place the links completely randomly; thus all nodes have the same chance of getting one […] If the network is large, despite the links' completely random placement, almost all nodes will have approximately the same number of links." (Albert-László Barabási, "Linked: How Everything Is Connected to Everything Else and What It Means for Business, Science, and Everyday Life", 2002)
"'There is an old debate', Erdos liked to say, 'about whether you create mathematics or just discover it. In other words, are the truths already there, even if we don't yet know them?' Erdos had a clear answer to this question: Mathematical truths are there among the list of absolute truths, and we just rediscover them. Random graph theory, so elegant and simple, seemed to him to belong to the eternal truths. Yet today we know that random networks played little role in assembling our universe. Instead, nature resorted to a few fundamental laws, which will be revealed in the coming chapters. Erdos himself created mathematical truths and an alternative view of our world by developing random graph theory." (Albert-László Barabási, "Linked: How Everything Is Connected to Everything Else and What It Means for Business, Science, and Everyday Life", 2002)
"In colloquial usage, chaos means a state of total disorder. In its technical sense, however, chaos refers to a state that only appears random, but is actually generated by nonrandom laws. As such, it occupies an unfamiliar middle ground between order and disorder. It looks erratic superficially, yet it contains cryptic patterns and is governed by rigid rules. It's predictable in the short run but unpredictable in the long run. And it never repeats itself: Its behavior is nonperiodic."
"Like regular networks, random ones are seductive idealizations. Theorists find them beguiling, not because of their verisimilitude, but because they're the easiest ones to analyze. [...] Random networks are small and poorly clustered; regular ones are big and highly clustered."
"In the telephone system a century ago, messages dispersed across the network in a pattern that mathematicians associate with randomness. But in the last decade, the flow of bits has become statistically more similar to the patterns found in self-organized systems. For one thing, the global network exhibits self-similarity, also known as a fractal pattern. We see this kind of fractal pattern in the way the jagged outline of tree branches look similar no matter whether we look at them up close or far away. Today messages disperse through the global telecommunications system in the fractal pattern of self-organization." (Kevin Kelly, "What Technology Wants", 2010)
"Although cascading failures may appear random and unpredictable, they follow reproducible laws that can be quantified and even predicted using the tools of network science. First, to avoid damaging cascades, we must understand the structure of the network on which the cascade propagates. Second, we must be able to model the dynamical processes taking place on these networks, like the flow of electricity. Finally, we need to uncover how the interplay between the network structure and dynamics affects the robustness of the whole system." (Albert-László Barabási, "Network Science", 2016)
No comments:
Post a Comment