(X2) N units 2 units N3 units Figure 5: A 2-N - 2 - 3 -...

 

Transcribed Image Text:

(X2) N units 2 units N3 units Figure 5: A 2-N - 2 - 3 - 1 model to make the binary classification for our dataset. (c) If Xavier initialization is used for the 2 N parameters in W, the N N param- eters in W2, the N2 N3 parameters in W3, and the 3 parameters in W4, what should be the four variances of the Gaussian distribution? (d) Build the 2 - N1 N2 N3 - 1 neural network of Figure 5 in the Jupyter note- book template received with the final exam PDF file. Set the activation functions to tanh for the hidden layers. The output activation should be a sigmoid. Write the model construction via TensorFlow/Keras. Save your Jupyter notebook before sending it back to the instructor by email. The training and the validation accuracy should be above 99%. (e) Replace the tanh activation of the hidden layers by ReLu. Try also sigmoids instead of tanh. What are the accuracy results after changing the activation functions? (X2) N units 2 units N3 units Figure 5: A 2-N - 2 - 3 - 1 model to make the binary classification for our dataset. (c) If Xavier initialization is used for the 2 N parameters in W, the N N param- eters in W2, the N2 N3 parameters in W3, and the 3 parameters in W4, what should be the four variances of the Gaussian distribution? (d) Build the 2 - N1 N2 N3 - 1 neural network of Figure 5 in the Jupyter note- book template received with the final exam PDF file. Set the activation functions to tanh for the hidden layers. The output activation should be a sigmoid. Write the model construction via TensorFlow/Keras. Save your Jupyter notebook before sending it back to the instructor by email. The training and the validation accuracy should be above 99%. (e) Replace the tanh activation of the hidden layers by ReLu. Try also sigmoids instead of tanh. What are the accuracy results after changing the activation functions?