project.

Learning Goal: I’m working on a data engineering exercise and need the explanation and answer to help me learn.

Q1)Read carefully Chapter 1 from “Gulli, A., Kapoor, A., Pal, S. (2019). Deep Learning with TensorFlow 2 and Keras: Regression, ConvNets, GANs, RNNs, NLP, and More with TensorFlow 2 and the Keras API, 2nd Edition. United Kingdom: Packt Publishing.”

Install TensorFlow by following instructions at tensorflow.org. Then, complete the tutorial at ‘TensorFlow 2 quickstart for beginners‘ by doing following:

a.Load a prebuilt dataset.

b.Build a neural network machine learning model that classifies images.

c.Train this neural network.

d.Evaluate the accuracy of the model.

For the submission, submit in the plain text

a)The commands you ran for the 4 steps above

b)The output of probability_model(x_test[:5])

See the rubric for grading.

Source code for Chapter 1

Q2)Read carefully Chapter 2 from “Gulli, A., Kapoor, A., Pal, S. (2019). Deep Learning with TensorFlow 2 and Keras: Regression, ConvNets, GANs, RNNs, NLP, and More with TensorFlow 2 and the Keras API, 2nd Edition. United Kingdom: Packt Publishing.” Then, complete the tutorial at ‘Basic regression: Predict fuel efficiency‘ by doing following:

a.Load ‘Auto MPG dataset’.

1. Normalize the data.
2. Find linear regression between milage and horsepower.
3. Plot the linear regression line and the data on figure.

For the submission, submit in the plain text

a)The commands you ran for the steps above.

b)The figure showing linear regression line and the data.

See the rubric for grading.

Source code for Chapter 2

Q3) See Chapter 8: Recurrent Neural Networks from “Gulli, A., Kapoor, A., Pal, S. (2019). Deep Learning with TensorFlow 2 and Keras: Regression, ConvNets, GANs, RNNs, NLP, and More with TensorFlow 2 and the Keras API, 2nd Edition. United Kingdom: Packt Publishing,” example on page 292: ‘Example ‒ One-to-Many – learning to generate text’. Then, based on the example,

A.import the necessary libraries and set up the constants

1. download and prepare the data for our network to consume
2. create the vocabulary
3. use the mapping dictionaries to convert the character sequence input into an integer sequence
4. define the network
5. define a loss function and compile the model
6. run the training and evaluation loop
7. do 50 epochs of training

For the submission:

1.submit the plain text code for the steps above

1. interpret the results/output of the example above

See the rubric for grading.

Source code for Chapter 8