Algorithm 1 Learning Parameters Update Mechanism

1: Input: $\left(x\left(i\right),y\left(i\right)\right)$ as training data; $\eta$ as learning rate; $\nabla E\left(\theta \right)$ is a gradient of loss (error) function $E\left(\theta \right)$ with respect to the $\theta$ parameter; momentum factor ($m$) 2: Output: For each training pair, update the learning parameters using the equation θ=θ-η.∇E(θ;x(i);y(i)) 3: if stopping condition is met 4: return parameter θ.