1.
W hich OCI services should you use to develop interactive Spark applications and
deploy Spark models?
✔️ Analytica Cloud
✔️ Data Flow
2. Which model artifacts can be shared from the model catalog?
✔️ Metadata
✔️ Hyperparameters
✔️ Metrics
3. Which option should be used to create a sophisticated autoscaling query
combining multiple metrics?
✔️ Custom scaling metric with MQL expressions
4. Which component of OCI Data Flow allows reusable Spark templates?
✔️ Applications
5. Fastest way to deploy and test LLM in a chatbot?
✔️ Using AI Quick Actions
6. What happens when a model deployment is deactivated?
✔️ The model’s HTTP endpoint becomes unavailable
✔️ Metadata is preserved
7. Which service provides free public datasets?
✔️ Oracle Open Data
8. Difference between a job and a job run?
✔️ A job is a template
✔️ A job run is a single execution
9. What verb should be used in the policy for view-only access?
✔️ Inspect
10.Where can you access or install the ADS SDK?
✔️ Conda environments in OCI Data Science
✔️ Python Package Index (PyPI)
11.First step after running Oracle Resource Manager stack?
✔️ Add users to the created groups
12.Which statements are true about OCI Open Data Service?
✔️ Users can contribute to repositories
✔️ Includes text and image data repositories
13.How to apply all recommended transformations in ADS?
✔️ auto_transform()
14.How to save results even after notebook deactivation?
✔️ Store results in the block storage
15.Key difference between PDP and ICE?
✔️ PDP provides feature-level insights
✔️ ICE provides sample-level insights
16.Purpose of continuous training in MLOps?
✔️ To retrain models for redeployment
17.Incorrect statement about autoscaling benefits?
✔️ Cost of deployment remains constant irrespective of resource utilization
18.Unsupported encryption algorithm in OCI Vault?
✔️ SHA-256
19.Which model is preferred for cancer detection?
✔️ Model 2 (high recall)
20.Why notebook can’t access Object Storage after IAM update?
✔️ Notebook session needs to be restarted
21.Which type of data is NOT available in Oracle Open Data?
✔️ Financial transaction data
22.Which two metrics are used to evaluate regression models in ADS?
✔️ Explained Variance Score
✔️ Mean Absolute Error
23.How can a Data Science Project be created in OCI?
✔️ Through Console UI
✔️ ADS SDK
24.How can you invoke a deployed LLM model?
✔️ API
✔️ CLI
25.Which two methods help verify model accuracy after policy change?
✔️ Drift monitoring
✔️ Validation with recent data
26.Which operator is best for redacting personal details?
✔️ PII Detection Operator
27.How to fetch data from Autonomous DB without wallet?
✔️ Provide hostname and port in connection parameters
28.Which three formats are supported by ADS for loading data?
✔️ Pandas DataFrame
✔️ JSON
✔️ XML
29.First step before training image classification model?
✔️ Label the images
30.Primary reason for feature scaling?
✔️ To bring features into the same scale
31.How is DAG built using Pipelines?
✔️ Using dependencies
32.Which two metrics are used for multiclass classification evaluation?
✔️ F1 Score
✔️ Recall
33.Default model parameter when unsure in Forecasting operator?
✔️ auto
34.Which interface to use for Python script to manage projects?
✔️ OCI SDK
35.Recommended environment for iterative PySpark development?
✔️ Notebook session
36.Best approach for hyperparameter tuning with known values?
✔️ Define a custom search space
37.Difference between one-hot and label encoding?
✔️ One-hot creates binary columns
✔️ Label assigns integers
38.Tool to correct class imbalance in ADS?
✔️ sample()
39.Which two statements are true about published Conda environments?
✔️ Curated by OCI
✔️ Configured using ads conda install
40.Correlation method for categorical variables in ADS?
✔️ Cramér’s V
41.Which file should be modified to include missing dependencies?
✔️ runtime.yaml
42.Correct sequence for deploying RAG application?
✔️ Initialize → Configure → Run → Check results
43.True statement about autoscaling configuration in Active state?
✔️ Must update one field at a time
44.Which interfaces can save model artifact to catalog?
✔️ OCI Console
✔️ CLI
45.Best way to find model framework details in catalog?
✔️ Model Taxonomy
46.Which step is automated by ADS during job run?
✔️ Saving outputs to Object Storage
47.Where are training job outputs stored?
✔️ OCI Object Storage
48.How to ensure same environment during deployment?
✔️ Use same Conda environment as training
49.How does AI Quick Actions ensure reliable responses?
✔️ By evaluating the model
50.Which activity is NOT part of ML lifecycle?
✔️ Database Management
51.Which component helps organize ML workflows in OCI?
✔️ Projects
52.Which file defines inference logic for model deployment?
✔️ score.py
53.Which stage helps identify data imbalance?
✔️ Data Exploration
54.Which statement is true about metric-based autoscaling?
✔️ Metrics are averaged across all instances
55.Which step is part of AutoML pipeline?
✔️ Feature Selection
56.Which function represents prediction error?
✔️ Cost function
57.What helps reduce frequent autoscaling adjustments?
✔️ Cooldown period
58.Which OCI service allows scalable Spark applications?
✔️ OCI Data Flow
59.Which ML library is used in Apache Spark?
✔️ MLlib
60.What should be examined first if regression predictions deviate?
✔️ Loss function
61.Best visualization for two continuous variables?
✔️ Scatterplot
62.Correct Git workflow order?
✔️ Install → Configure SSH → Create repo → Commit → Push
63.Correct sequence for using Data Science Operators?
✔️ Install Conda → Initialize → Configure → Run → Check results
64.Which file types can be annotated in Data Labeling?
✔️ Images
✔️ Typed documents
✔️ Purchase orders
65.Class types for document annotation?
✔️ Single label
✔️ Multiple label
✔️ Key Value
66.Correct sequence for forecasting model steps?
✔️ Prepare → Verify → Save → Deploy → Predict
67.Purpose of dynamic group in OCI?
✔️ Manage access for resources like notebook sessions
68.Method to preview dataset in ADS?
✔️ show_in_notebook()
69.Primary advantage of Conda environments?
✔️ Isolated software configurations
70.Technique to explain specific flagged transaction?
✔️ Local Explanation
71.Which visualization does ADS suggest for monthly charges?
✔️ Scatterplot
72.Which two Git operations are affected by slow internet?
✔️ Push
✔️ Pull
73.Which two IAM setups can fix a 404 error when saving a model?
✔️ Dynamic group with matching rules
✔️ Policy granting manage access to model catalog
74.Autoscaling policy for CPU utilization?
✔️ Predefined metric
75.How to install Python packages during notebook creation?
✔️ Use runtime configuration
76.Primary goal of loss function?
✔️ Quantify difference between prediction and target
77.Best way to fix disk space error in job run?
✔️ Edit job to increase block storage and rerun
78.Best description of Data Science Jobs?
✔️ Define and run repeatable tasks on managed infrastructure
79.What happens when job ends?
✔️ Infrastructure is automatically deprovisioned
80.Best explanation technique for correlated features?
✔️ Accumulated Local Effects (ALE)
81.Best way to store external DB credentials?
✔️ OCI Vault
82.How to ensure data processing before model training in pipeline?
✔️ Set step dependencies
83.Core idea behind LIME?
✔️ Local model-agnostic explanation
84.Feature to store and version models?
✔️ Model Catalog
85.Which model is preferred for cancer detection?
✔️ Model 2 (high recall)
86.Purpose of score.py file?
✔️ Define inference logic
87.Which stage helps identify imbalances in data?
✔️ Data Exploration
88.True statement about metric-based autoscaling?
✔️ Metrics are averaged across all instances
89.Which step is part of AutoML pipeline?
✔️ Feature Selection
90.Function representing prediction error?
✔️ Cost function
91.What helps reduce frequent autoscaling adjustments?
✔️ Cooldown period
92.OCI service for scalable Spark applications?
✔️ OCI Data Flow
93.ML library used in Apache Spark?
✔️ MLlib
94.What to examine first if regression predictions deviate?
✔️ Loss function
95.Best visualization for two continuous variables?
✔️ Scatterplot
96.Correct Git workflow order?
✔️ Install → Configure SSH → Create repo → Commit → Push
97.Correct sequence for using Data Science Operators?
✔️ Install Conda → Initialize → Configure → Run → Check results
98.Which file types can be annotated in Data Labeling?
✔️ Images
✔️ Typed documents
✔️ Purchase orders
99.Class types for document annotation?
✔️ Single label
✔️ Multiple label
✔️ Key Value
100. Correct sequence for forecasting model steps?
✔️ Prepare → Verify → Save → Deploy → Predict
101. Purpose of dynamic group in OCI?
✔️ Manage access for resources like notebook sessions
102. Method to preview dataset in ADS?
✔️ show_in_notebook()
103. Primary advantage of Conda environments?
✔️ Isolated software configurations
104. Technique to explain specific flagged transaction?
✔️ Local Explanation
105. Which visualization does ADS suggest for monthly charges?
✔️ Scatterplot
106. Which two Git operations are affected by slow internet?
✔️ Push
✔️ Pull
107. Two IAM setups to fix 404 error when saving model?
✔️ Dynamic group with matching rules
✔️ Policy granting manage access to model catalog
108. Autoscaling policy for CPU utilization?
✔️ Predefined metric
109. How to install Python packages during notebook creation?
✔️ Use runtime configuration
110. Primary goal of loss function?
✔️ Quantify difference between prediction and target
111. Best way to fix disk space error in job run?
✔️ Edit job to increase block storage and rerun
112. Best description of Data Science Jobs?
✔️ Define and run repeatable tasks on managed infrastructure
113. hat happens when job ends?
W
✔️ Infrastructure is automatically deprovisioned
114. Best explanation technique for correlated features?
✔️ Accumulated Local Effects (ALE)
115. Best way to store external DB credentials?
✔️ OCI Vault
116. How to ensure data processing before model training in pipeline?
✔️ Set step dependencies
117. Core idea behind LIME?
✔️ Local model-agnostic explanation
118. Feature to store and version models?
✔️ Model Catalog
119. True statement about autoscaling configuration in Active state?
✔️ Must update one field at a time
120. Which interfaces can save model artifact to catalog?
✔️ OCI Console
✔️ CLI
121. Best way to find model framework details in catalog?
✔️ Model Taxonomy
✔️ Provenance details
122. Which step is automated by ADS during job run?
✔️ Saving outputs to Object Storage
123. Where are training job outputs stored?
✔️ OCI Object Storage
124. How to ensure same environment during deployment?
✔️ Use same Conda environment as training
125. How does AI Quick Actions ensure reliable responses?
✔️ By evaluating the model
126. Which activity is NOT part of ML lifecycle?
✔️ Database Management
127. Invalid notebook session approach?
✔️ Manually downloading data and uploading to block volume
128. Purpose of score.py file?
✔️ Define inference logic
129. Where are OCI secrets stored?
✔️ OCI Vault
130. Invalid OCI Data Science policy?
✔️ Allow dynamic-group to manage data-science-family in compartment
131. Where to specify Conda environment in model artifact?
✔️ runtime.yaml
132. How to avoid model deployment hitting utilization limit?
✔️ Add more instances or increase VM size
133. How to calculate optimal load balancing bandwidth?
✔️ Use payload size × requests/sec × buffer
134. est False Alarm Probability setting for anomaly detection?
B
✔️
Low
135. Which Conda environment task requires activation?
✔️
Installing packages
136. Best way to identify optimal compute shape?
✔️
Start small, monitor metrics, scale gradually
137. CLI command to share Conda environment?
✔️
ads conda publish
138. Open model format for cross-platform ML?
✔️
ONNX
139. Best way to pass parameters to job runs?
✔️
Environment variables
✔️
Command-line arguments
140. What happens when notebook session is deactivated?
✔️
Compute stops
✔️
Block volume persists
