Support vector machine
<p>Feb 06, 2011 01:54 PM
 
supervised learning
 
used for classification and regression analysis.
 
Inventor  Vladimir Vapnik.
 
SVM predicts whether a new example falls into one category or the other.
 
maps the points into a new space such that the cldar gap between them is as wide as possible.
 
SVM wants to achieve the hyperplane that has the largest distance to the nearest training data points of any class
(the "functional margin").
 
The original problem may not be linearly separable.
 
SVM uses a mapping into a larger space so that corss porducts are easier to compute.
 
The cross products are defined in terms of a KERNEL FUNCTION K(x,y).
 
 
 
 </p>