Consider a discrete time system described asi y(t + 1) = -0.5y(t) yt - 1) + 0.5 ut) (I) A Recurrent network with a linear activation function is chosen to identify the system dynamics. The response of the network is given asi y(t + 1) = wy(t) + wzyt 1) + W3 u(t) where w1, W2 and wz are the unknown weights. Derive weight update law for the three weights using RTRL (Real time recurrent learning) so that the parameters converge to the actual system parameters associated with the dynamics given by (I). The cost function is to be taken as E(t+1) = { v(t + 1) y(t+1)]2 Also draw the Recurrent network that can learn the dynamics given by (I). Consider a discrete time system described asi y(t + 1) = -0.5y(t) yt - 1) + 0.5 ut) (I) A Recurrent network with a linear activation function is chosen to identify the system dynamics. The response of the network is given asi y(t + 1) = wy(t) + wzyt 1) + W3 u(t) where w1, W2 and wz are the unknown weights. Derive weight update law for the three weights using RTRL (Real time recurrent learning) so that the parameters converge to the actual system parameters associated with the dynamics given by (I). The cost function is to be taken as E(t+1) = { v(t + 1) y(t+1)]2 Also draw the Recurrent network that can learn the dynamics given by (I)