Inference Problem
At test time, only a sample is known, and a prediction function maps it to a predicted label from the label space . For structured SVMs, given the vector obtained from training, the prediction function is the following.
Therefore, the maximizer over the label space is the predicted label. Solving for this maximizer is the so-called inference problem and similar to making a maximum a-posteriori (MAP) prediction in probabilistic models. Depending on the structure of the function, solving for the maximizer can be a hard problem.
Read more about this topic: Structured SVM
Famous quotes containing the words inference and/or problem:
“I shouldn’t want you to be surprised, or to draw any particular inference from my making speeches, or not making speeches, out there. I don’t recall any candidate for President that ever injured himself very much by not talking.”
—Calvin Coolidge (1872–1933)
“The great problem of American life [is] the riddle of authority: the difficulty of finding a way, within a liberal and individualistic social order, of living in harmonious and consecrated submission to something larger than oneself.... A yearning for self-transcendence and submission to authority [is] as deeply rooted as the lure of individual liberation.”
—Wilfred M. McClay, educator, author. The Masterless: Self and Society in Modern America, p. 4, University of North Carolina Press (1994)