In pattern recognition a feature space is an abstract space where each pattern sample is represented as a point in n-dimensional space. Its dimension is determined by the number of features used to describe the patterns. Similar samples are grouped together, which allows the use of density estimation for finding patterns.
The concept is a most used one in classification techniques like k nearest neighbors or support vector machines.
Famous quotes containing the words feature and/or space:
“Blunders are an inescapable feature of war, because choice in military affairs lies generally between the bad and the worse.”
—Allan Massie (b. 1938)
“Here were poor streets where faded gentility essayed with scanty space and shipwrecked means to make its last feeble stand, but tax-gatherer and creditor came there as elsewhere, and the poverty that yet faintly struggled was hardly less squalid and manifest than that which had long ago submitted and given up the game.”
—Charles Dickens (1812–1870)