Train the neural algorithm usinMetwork in Figure Q.4(i) using the sequential gradient descent g the data in Table Q.4(i). Fill Table Q.4(ii) and Table Q.4(iii) for onTM epoch. Use learning rate, a₁ = 0.1 f 0.1 for Table Q.4(ii) and learning rate. Table Q.4(iii). All the Table Q.4(ii) Tabghts and biases for both tables are Q.4(iii). The neurons in the g rate, d₂ = 0.9 for layer and the output layer use linear activation function while the neur in Unitialized as shown in in the hidden layer use SigmUTN activation function, f(b) = 1 1+e-b Show all your working steps. wl11 w211 X 0.4 UTM Figure Q.4(i) w210 y Epoch 0 Input, X TM 1 1 Epoch 0 1 2 Table Q.4(i) Point, i Input, X ITM 2 Target, Y - 1 1 2 Table Q.4(ii) STMUT Input, X w111 Target, Y 1 TM TM 2 0.3 10 0.5 w211 0.4 Error UTM TM TM Table Q.4(iii) Target, Y UTM TM - 1 w111 w210 w211 Error 0.3 0.5 0.4 2 2

Train the neural algorithm usinMetwork in Figure Q.4(i) using the sequential gradient descent g the data in Table Q.4(i). Fill Table Q.4(ii) and Table Q.4(iii) for onTM epoch. Use learning rate, a₁ = 0.1 f 0.1 for Table Q.4(ii) and learning rate. Table Q.4(iii). All the Table Q.4(ii) Tabghts and biases for both tables are Q.4(iii). The neurons in the g rate, d₂ = 0.9 for layer and the output layer use linear activation function while the neur in Unitialized as shown in in the hidden layer use SigmUTN activation function, f(b) = 1 1+e-b Show all your working steps. wl11 w211 X 0.4 UTM Figure Q.4(i) w210 y Epoch 0 Input, X TM 1 1 Epoch 0 1 2 Table Q.4(i) Point, i Input, X ITM 2 Target, Y - 1 1 2 Table Q.4(ii) STMUT Input, X w111 Target, Y 1 TM TM 2 0.3 10 0.5 w211 0.4 Error UTM TM TM Table Q.4(iii) Target, Y UTM TM - 1 w111 w210 w211 Error 0.3 0.5 0.4 2 2