Algorithm SDM (c, b, w)

θ1: inference error

Tmax1: the maximum number of learning time

n: the number of rules

input: current parameters

output: parameters c, b, and w after learning

Steps A3 to A8 of Algorithm A are performed.
