This page is a one-stop solution for any information you may require for SAS Viya Natural Language Processing and Computer Vision (A00-408) Certification exam. The SAS A00-408 Exam Summary, Syllabus Topics and Sample Questions provide the base for the actual SAS Viya Natural Language Processing and Computer Vision exam preparation, we have designed these resources to help you get ready to take your dream exam.
The SAS Viya Natural Language Processing and Computer Vision credential is globally recognized for validating SAS Viya Natural Language Processing and Computer Vision knowledge. With the SAS Viya Natural Language Processing and Computer Vision Certification credential, you stand out in a crowd and prove that you have the SAS Viya Natural Language Processing and Computer Vision knowledge to make a difference within your organization. The SAS Viya Natural Language Processing and Computer Vision Certification (A00-408) exam will test the candidate's knowledge on following areas.
SAS A00-408 Exam Summary:
| Exam Name | SAS Viya Natural Language Processing and Computer Vision |
| Exam Code | A00-408 |
| Exam Duration | 110 minutes |
| Exam Questions | 60 |
| Passing Score | 70% |
| Exam Price | $180 (USD) |
| Books |
SAS Visual Text Analytics in SAS Viya Deep Learning Using SAS Software |
| Exam Registration | Pearson VUE |
| Sample Questions | SAS Viya Natural Language Processing and Computer Vision Certification Sample Question |
| Practice Exam | SAS Viya Natural Language Processing and Computer Vision Certification Practice Exam |
SAS A00-408 Exam Topics:
| Objective | Details |
|---|---|
Loading and Exploring Data (18-22%) |
|
| Import documents for analysis |
- Convert documents for analysis. - Explore and prepare a document. - Troubleshoot Language encoding issues (ASCII, UTF-8, etc.). - Given a scenario, ensure minimal loss of information when converting documents from proprietary formats to SAS supported formats. |
| Create and explore a project in SAS Visual Text Analytics |
- Identify the SAS Visual Text Analytics default pipeline. - Explore the Documents Table. - Identify and define key features of the term table. - Given a scenario, appropriately assign text and category roles. - Export score code to score new data sets. |
| Load and prepare image data |
- Load labeled image data (labelLevels, loadImages). - Augment image data. - Prepare data for modeling. |
Identifying Text Patterns Using Natural Language Processing Techniques (40-45%) |
|
| Use the Concepts and Text Parsing Nodes to extract Terms and Concepts |
- Use lists to include or exclude or combine terms (i.e. start, stop, synonym). - Explain why Concepts are useful. - Explain predefined Concepts. - Define custom Concepts for a project. - Modify the Term Table and explain the impact on the pipeline. - Explain the impact of concepts on the pipeline. - View document matches and similarity scores. - Explore the term map (identify various components). |
| Write Concept Rules |
- Given a scenario, use LITI to write a rule to achieve a goal (i.e. CATEGORY, CLASSIFIER, CONCEPT, C_CONCEPT, CONCEPT_RULE, NO_BREAK, PREDICATE_RULE, REGEX, etc.) - Given a LITI rule, explain the how it influences scoring documents. - Given a LITI rule, explain the how it impacts the term table. - Identify and correct common syntax errors |
| Use the Topics Node to extract machine-generated topics |
- Given a scenario, appropriately adjust term density. - Given a scenario, appropriately adjust document density. - Promote a topic to a category. - Split and merge topics. - Edit Topic Properties. - Create Custom Topics. |
| Use rules to identify documents belonging to specific categories |
- Analyze categorization results (F-Measure, precision, recall, misclassification). - Edit and enhance predefined rules with defined concepts. - Explain Categories Results (diagnostic counts, diagnostic metrics, categories score code). - View Document matches and sentiment score. - Explain sentiment level scoring. |
| Write Category rules |
- Given a scenario, create and run appropriate Boolean rules to achieve a goal. - Given a rule, explain how it impacts document categorization. |
| Use a Recurrent Neural Network (RNN) to recognize patterns |
- Build a Basic RNN. - Build a Bi-directional RNN. - Build a Specialized (GRU, LSTM) RNN. |
Identifying Image Patterns Using Computer Vision Techniques (35-40%) |
|
| Use convolutional layers in a Convolutional Neural Network (CNN) |
- Explain the use of kernel filters in a CNN. - Explain and calculate feature maps in a CNN (i.e. size). - Detail equivariance to translation. - Define hyperparameters (width, height and stride). - Detail number of weights. |
| Use padding in a Convolutional Neural Network |
- Detail the impact of padding on the feature map size. - Use padding to accommodate skip-layer connections. - Given a scenario, use padding to accomplish a goal. |
| Use pooling in a Convolutional Neural Network |
- Detail the impact of pooling on the invariance of the CNN. - Define summary functions used in pooling layers. - Explain the use of filters in a CNN. - Given a scenario, determine if using pooling is appropriate. |
| Use fully connected layers in a Convolutional Neural Network |
- Given specific action calls, define number of parameters (trainable, estimated, etc.). - Order FC layers correctly in building a CNN. - Define activation functions used in Fc layers. |
| Use output layers in a Convolutional Neural Network |
- Specify activation function for the output layer. - Define types of error functions. |
| Tune the Hyperparameters of a Convolutional Neural Network | - Tune a deep learning model using the Hyperband method. |
| Score new image data |
- Use trained weights to score new image data. - Explain the relevance of batch size in scoring new image data. |
| Explain the impact of various architectural designs |
- Use residual connections. - Use concatenation connections. - Define requirements for skip layer connections. - Use one-by-one convolutions. - Use Spatial Exploration techniques. - Define blocks. - Use cardinality techniques in the network structure. |
| Use regularization techniques |
- Explain batch normalization. - Use batch normalization to improve model generalization and learning. - Explain dropout. - Use dropout to improve model generalization. |
The SAS has created this credential to assess the knowledge and understanding of a candidate in the area as above via the certification exam. The SAS Viya Natural Language Processing and Computer Vision (A00-408) Certification exam contains a high value in the market being the brand value of the SAS attached with it. It is highly recommended to a candidate to do a thorough study and also get a hand full of the practice to clear SAS Viya Natural Language Processing and Computer Vision exam without any hiccups.