Model algorithm

Randomly initialize the coefficients $w_{ij}$

Let $x\left(n\right)=\left\{x_1\left(n\right),\cdots ,x_p\left(n\right)\right\}$ , the entrance

Let $d\left(n\right)=\left\{d_1\left(n\right),\cdots ,d_m\left(n\right)\right\}$ , the exit

$\mu$ is a positive constant

$a_i$ is the activation value of the neuron $N_i$

beginning

Repeat:

Take a couple (x, d) in S

Output Error Calculation

Error back-propagation to inputs

Calculate network output for input x $y^{*}$

If $d\ne y^{*}$

Weight update:#Math_30#

$w_{ij}=w_{ij}+\mu \ast \left(a_i\ast a_j\right)$

End if

end repeat

Let D be the test basis

Application of Bayes’ Theorem

End