Linear

Linear

Linear kernel: k(x,y) = xᵀy + c. Parameters: [c] (default c = 0). nbParameters = 1. Not scalable.

Constructor

new Linear(copt)

Source:
Parameters:
Name Type Attributes Default Description
c number <optional>
 0

Classes

Linear

Members

c :number

Source:
Type:
  • number

Methods

gradient(_x1, _x2) → {Array.<number>}

Description:

  • gradient = [grad_c] = [1].

Source:
Parameters:
Name Type Description
_x1 Array.<number>
_x2 Array.<number>
Returns:
Type
Array.<number>

heuristicFit()

Description:

  • Default no-op heuristic fit.

Source:

kernel(x1, x2) → {number}

Source:
Parameters:
Name Type Description
x1 Array.<number>
x2 Array.<number>
Returns:
Type
number

setParameters(parameters)

Source:
Parameters:
Name Type Description
parameters Array.<number>