Glossary

action potential Also called a "spike," the all-or-none response elicited by a neuron, the frequency of which depends upon the total input to the neuron. A sequence of action potentials is called a "spike train.''

binding problem The problem of how activities of neurons, which are distributed across the brain, are linked to represent the same object.

compartmental model A biophysical model of a neuron based on a series of circuit elements representing passive and active electrical properties of cell membranes.

credit-assignment problem The problem of determining the contribution of each neuron and connection weight to the output of the network.

emergent computation Computation resulting from the interactions of many neurons within a network.

objective function The function that is optimized in training artificial neural networks and usually includes a term related to the error of the network.

supervised training A training procedure that is based on a set of exemplars composed of sample input data and the correct or desired output—"truth data.'' The error between the actual and desired output is directly incorporated into the objective function.

synapse From the Greek word "to clasp,'' the location of electrochemical communication between connected neurons. In artificial neural networks synapses are represented by connection weights.

unsupervised training A training procedure that does not rely on labeled "truth" data. Often used in data clusters and the construction of associative memories.

Confidence and Social Supremacy

Confidence and Social Supremacy

Surefire Ways To Build Up Your Confidence As Well As Be A Great Networker. This Book Is One Of The Most Valuable Resources In The World When It Comes To Getting Serious Results In Building Confidence.

Get My Free Ebook


Post a comment