Preview Quiz.md

Instructions

This is quiz answer sheet only. In order to answer it you might need to do some code exercises provided in quiz guideline.

1. GAN is a generative model. Please select all box that represents a generative models
• [ ] Naive Bayes
• [ ] Logistic Regression
• [ ] Support Vector Machine
• [ ] K-Nearest Neighbor
2. What does adversarial refers to in GAN ?
• [ ] Loss function
• [ ] Optimizer
• [ ] Input
• [ ] Output
3. Select all correct boxes about Generator and Discriminator
• [ ] Generator takes input from Discriminator
• [ ] Discriminator tries to maximize the loss function
• [ ] Discriminator tries to generate a random noise
• [ ] GAN input is random vector
4. To ensure you have correctly load the data, what is the shape of X_train and X_test respectively?
• [ ] (60000, 28, 28, 1) and (10000, 28, 28, 1)
• [ ] (50000, 28, 28, 1) and (5000, 28, 28, 1)
• [ ] (60000, 32, 32, 1) and (10000, 32, 32, 1)
• [ ] (50000, 32, 32, 1) and (5000, 32, 32, 1)
5. While loading the Mnist dataset, in line 4, what does `X_train = (X_train.astype(np.float32) - 127.5)/127.5` means ?
• [ ] Rescale the data into within the interval of -1 to 1
• [ ] Rescale the data into within the interval of 0 to 1
• [ ] Denormalize data
• [ ] Reshape data
6. If you follow the quiz guide, how many params does the generator_model has ?
• [ ] 6,765,313
• [ ] 6,763,777
• [ ] 6,751,233
• [ ] 12,544
7. If you change the last convolution layer's filter shape into 7x7, what will be the final total parameters ?
• [ ] 6,765,313
• [ ] 6,763,777
• [ ] 6,751,233
• [ ] 12,544
8. What does Flatten() layer means in the Discriminator?
• [ ] To make the data standardized
• [ ] To make the data mean = 0
• [ ] To reshape the data into one dimensional vector
• [ ] To compile the previous layers
9. What is the differences between images that are not checked by discriminator (directly generated from generator), and images that previously checked by discriminator?
• [ ] The unchecked images are smaller in size
• [ ] The checked images are usually better
• [ ] The unchecked images are usually better
• [ ] The checked images are smaller in size
10. What do you think about losses over time while training ?
• [ ] It will be minimized overtime
• [ ] It will be maximized overtime
• [ ] It will just oscilate overtime
Quiz
You need to score 8 out of a possible 10 to earn a badge.
You have 2 attempts. Only your highest score will be taken into account.
• ###### Quiz 1

• GAN is a generative model. Please select all box that represents a generative models
• Question worth 1 point

• ###### Quiz 2

• What does adversarial refers to in GAN ?
• Question worth 1 point

• ###### Quiz 3

• Select all correct boxes about Generator and Discriminator
• Question worth 1 point

• ###### Quiz 4

• To ensure you have correctly load the data, what is the shape of X_train and X_test respectively?
• Question worth 1 point

• ###### Quiz 5

• While loading the Mnist dataset, in line 4, what does `X_train = (X_train.astype(np.float32) - 127.5)/127.5` means ?
• Question worth 1 point

• ###### Quiz 6

• If you follow the quiz guide, how many params does the generator_model has ?
• Question worth 1 point

• ###### Quiz 7

• If you change the last convolution layer's filter shape into 7x7, what will be the final total parameters ?
• Question worth 1 point

• ###### Quiz 8

• What does Flatten() layer means in the Discriminator?
• Question worth 1 point

• ###### Quiz 9

• What is the differences between images that are not checked by discriminator (directly generated from generator), and images that previously checked by discriminator?
• Question worth 1 point

• ###### Quiz 10

• What do you think about losses over time while training ?
• Question worth 1 point

Recipients 2