In game theory, a pair of strategies for a game such that
neither player can improve his outcome by changing his strategy. A
Nash equilibrium sometimes takes the form of a saddle structure.
Under some cases, when a strategy is at a Nash equilibrium with
itself, the strategy resembles an evolutionary stable strategy.
Any of the standard counting numbers; a positive integer.
The natural filtering process by which individuals with higher
fitness are more likely to reproduce than individuals with lower
A synthesis of Darwinism with the mechanisms of genetics; the
idea that adaptation equals a combination of variation,
heredity, and selection. See also evolution, inheritable,
and natural selection.
The weighted sum of incoming signals into a neuron plus a
Neural Network (NN)
A network of neurons that are connected through synapses or
weights. In this book, the term is used almost exclusively to
denote an artificial neural network and not the real thing. Each
neuron performs a simple calculation that is a function of the
activations of the neurons that are connected to it. Through
feedback mechanisms and/or the nonlinear output
response of neurons, the network as a whole is capable of performing
extremely complicated tasks, including universal computation and
universal approximation. Three different classes of neural
networks are feedforward, feedback, and recurrent neural
networks, which differ in the degree and type of connectivity
that they possess.
A simple computational unit that performs a weighted sum on incoming
signals, adds a threshold or bias term to this value to yield a
net input, and maps this last value through an activation
function to compute its own activation. Some neurons, such as
those found in feedback or Hopfield networks, will
retain a portion of their previous activation.
An iterative method for finding 0 values of a function.
A way for an animal to make a living in an ecosystem.
No Free Lunch (NFL)
A theorem that states that in the worst case, and averaged over
an infinite number of search spaces, all search methods
perform equally well. More than being a condemnation of any search
method, the NFL theorem actually hints that most naturally occurring
search spaces are, in fact, not random.
A function that is not linear. Most things in nature are
nonlinear. This means that in a very real way, the whole is at least
different from the sum of the parts. See also holism.
Not Recursively Enumerable (not-RE)
An infinite set that cannot be recursively enumerated.
Sets of this type that are also not co-recursively enumerable
are effectively random.
Nondeterministic polynomial time problems; a class of computational
problems that may or may not be solvable in polynomial time but
are expressed in such a way that candidate solutions can be tested for
correctness in polynomial time. See also time complexity and
A problem type in which any instance of any other NP class
problem can be translated to in polynomial time. This means that
if a fast
algorithm exists for an NP-complete problem, then any problem
that is in NP can be solved with the same algorithm.
A solution to a problem that is calculated through a simulation.
For example, solving the Three Body Problem is not possible in
the worst case; however, with the differential equations that
describe the motions of three bodies in space, one could simulate
their movements by simulating each time step. Nevertheless, numerical
solutions are usually error-prone due to sensitivity and,
therefore, can be used to estimate the future for only relatively
short time spans, in the worst case.