Pass A00-408 Exam to Get SAS Institute Certification

The A00-408 exam is hot, and passing it requires a deep understanding of SAS Institute solutions. Practicing with SAS Institute A00-408 dumps questions can help you reinforce your knowledge and increase your chances of passing the exam. A00-408 dumps are available to help you prepare for the A00-408 exam. Using A00-408 exam dumps questions is one effective way to supplement your study plan and increase your chances of success on exam day. Test free SAS Certified Specialist: Natural Language Processing and Computer Vision Using SAS Viya A00-408 exam dumps below.

Page 1 of 3

1. The output for a particular neuron in the hidden layer is -0.001.

What activation function is used in this hidden layer?

RelU

TanH

Sigmoid

Softmax

 Loading...

2. 1.Given the code specifying the two initial layers of a convolutional network:

deepLearn.addLayer / layer={type="INPUT"

nchannels=1 width=10 height=10}

modelTable={name="simple"} name="data";

deepLearn.addLayer / layer={type="CONVO"

nFilters=10 width=5 height=5 includeBias=True}

modelTable={name="simple"} name="conv1"

srcLayers={"data"};

deepLearn.addLayer / layer={type='POOL'

width=2 height=2 stride=2 pool='max' }

modelTable={name="simple"} name="pool1"

srcLayers={"conv1"}

How many parameters are in the model?

259

260

250

274

 Loading...

3. Given the following code to add layers to a convolutional network, fill in the blank with the appropriate activation function that ensures batch normalization is computed correctly.

deepLearn.addLayer / model=’simple’ name=’conv’

layer={type=”CONVO” nFilters=10

width=3 height=3

act=”

 Loading...

4. Which statement is FALSE regarding the loadImages action?

All images are represented in a single table after loading.

Images can be 3-dimensional.

Images from multiple sources can be loaded in a single call.

Encoded images must be decoded prior to loading.

 Loading...

5. Enter the keyword for the blank line below that will return the string EM if both the string EM and the words Expectation Maximization are in a document. CLASSIFIER: [ _______ =EM]: Expectation Maximization

 Loading...

6. Which sentences would return a match for this Categories rule: (AND, (OR,"hotel", "motel", "inn"), (OR,"restaurant", "pool", "gift shop"))? (Choose 2)

The hotel had an indoor and outdoor pool.

As far as motels go, this one had a great restaurant.

I've been to many hotels, motels and inns.

Don't buy souvenirs from a gift shop in a hotel.

 Loading...

7. Which of the following is a best practice before building deep learning models after loading images into a CAS Table using the loadImages action?

Resize the images so they are all 32 x 32 pixels.

Convert the images into a flattened format.

Remove color channels from the image so they are greyscale.

Shuffle the images so that they are randomly ordered.

 Loading...

8. Which of the following steps is required before training a recurrent neural network to classify text documents?

Calculate or download pretrained word embeddings.

Parse the terms in the document corpus to perform stemming.

Clean the text data by removing non-letters and non-numbers.

Remove stop words from the document corpus.

 Loading...

9. The output CAS table generated by running the loadImages action contains the binary images in which of the following columns?

_path_

_jpg_

_binary_

_image_

 Loading...

10. What is the meaning of the pop-up box in the exhibit?

The color of the node shows the distance from the central term.

There are 52 documents that have all of the terms in the box excluding the last one.

The box shows the order that the terms appear in the selected documents.

The box shows the terms that predict the occurrence of the central term.

 Loading...

Page 2 of 3

11. What is a computational benefit of using LSTM and GRU layers in recurrent neural networks?

LSTM and GRU layers reduce the number of calculations needed in each training iteration.

LTSM and GRU layers enable parallelization of the model training across multiple machines.

LTSM and GRU layers reduce the overall number of weights in the model.

LSTM and GRU layers reduce the vanishing gradient problem.

 Loading...

12. What can the Sandbox be used for in Visual Text Analytics nodes?

Evaluate results of user topics before activating them in production.

Evaluate scoring code of the concepts node before activating it in production.

Evaluate rule syntax of a categories rule before activating it in a production rule.

Evaluate rule syntax of a concepts rule before activating it in a production rule.

 Loading...

13. Which rule types support the _c{} modifier? (Choose 2)

CLASSIFIER

CONCEPT_RULE

PREDICATE_RULE

C_CONCEPT

 Loading...

14. After a visual text analytics project is created, where on the data tab could you click if you wanted to see the raw text and data values?





A)





B)





C)





D)

Option A

Option B

Option C

Option D

 Loading...

15. Which of the following are capabilities of the image action set? (Choose 2)

It can display images once they've been uploaded into a CAS Table.

It can display all CAS Tables that contain images that are in memory.

It can mutate images via rotations, flips, color jitters, and more.

It can train deep learning models using CAS Tables that contain images.

 Loading...

16. Given the code specifying the output layer in a convolutional network designed for image classification:

deepLearn.addLayer / layer={type=”OUTPUT”

act=”SOFTMAX”}

modelTable={name=”simple”}

name=”outlayer”

srcLayers={“fc”};

What is the error function used by this image classification model?

CTC

POISSON

NORMAL

ENTROPY

 Loading...

17. Which of the following is an advantage of using a rectified linear transformation (ReLU) activation function in a convolutional network?

The output of the ReLU activation function is bounded above and thus ReLU never diverges

The gradient of the ReLU activation function is always 0 or 1 and is thus computationally easy to calculate

The ReLU activation function has no saturation region and thus training never stalls

The ReLU activation function allows parallelization of the training process thus speeding training

 Loading...

18. Which statement is true regarding a SAS Visual Text Analytics project?

There are only three roles the user can assign for variables: Text, Category, and none.

A variable with a role of Category is required to define a project.

Text and Category variables are the only variable types that can be used as Display variables.

There is only a single Visual Text Analytics pipeline available to users.

 Loading...

19. Which statements apply to importing documents to SAS Visual Text Analytics? (Choose 3)

You can import documents directly from a folder.

You can import documents directly for a SAS database.

You can import documents directly from a spreadsheet.

You can import documents directly from a webpage.

 Loading...

20. Which statement is TRUE regarding encoding?

Latin encoding is an extension of ASCII encoding with specific extensions.

Asian languages (i.e. Japanese) are typically encoded in ASCI

UTF-8 is a subset of ASCII encoding.

A UTF-8 character is 4 bytes long.

 Loading...

Page 3 of 3

21. Given the code to add a convolution layer:

AddLayer/model='mymodel' name="conv2"

layer={type='convolution' nFilters=32 width=1 height=1 stride=1}

includeBias=True

srcLayers={"conv1"};

Assume the output data dimension of the “conv1” layer is 100*100*10 (width*height*depth)

What is the number of trainable parameters for the layer “conv2”?

320

32

42

352

 Loading...

 Loading...