Machine Learning Algorithms (7) Support Vector Machines

The more fault-tolerant the better

b is the deflection direction of the plane, w is the normal vector of the plane, and the mapping of x to the plane:

The first thing is to find the point with the smallest distance from the dividing line; then what kind of w and b are obtained, so that such a point has the largest value from the dividing line.

After scaling: ; Take it as min again , that is, take yi*(w^T*Q(xi) + b) = 1 => 

 

Guess you like

Origin http://43.154.161.224:23101/article/api/json?id=325043088&siteId=291194637