Feature Selection
Feature selection approaches try to find a subset of the original variables (also called features or attributes). Two strategies are filter (e.g. information gain) and wrapper (e.g. search guided by the accuracy) approaches. See also combinatorial optimization problems.
In some cases, data analysis such as regression or classification can be done in the reduced space more accurately than in the original space.
Read more about this topic: Dimension Reduction
Famous quotes containing the words feature and/or selection:
“When delicate and feeling souls are separated, there is not a feature in the sky, not a movement of the elements, not an aspiration of the breeze, but hints some cause for a lover’s apprehension.”
—Richard Brinsley Sheridan (1751–1816)
“Judge Ginsburg’s selection should be a model—chosen on merit and not ideology, despite some naysaying, with little advance publicity. Her treatment could begin to overturn a terrible precedent: that is, that the most terrifying sentence among the accomplished in America has become, “Honey—the White House is on the phone.””
—Anna Quindlen (b. 1952)