



































































Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
This practice exam is a forward-looking evaluation tool covering emerging certification areas in HR, management, technology, digital transformation, workforce analytics, and innovation leadership. It introduces candidates to next-generation professional competencies, new-age frameworks, and futuristic skill domains needed to stay competitive in evolving industries.
Typology: Exams
1 / 75
This page cannot be seen from the preview
Don't miss anything!




































































Question 1. Which learning paradigm trains a model using input–output pairs where the correct answer is known during training? A) Unsupervised Learning B) Reinforcement Learning C) Supervised Learning D) Self-Supervised Learning Answer: C Explanation: Supervised learning uses labeled data (input-output pairs) to teach the model the mapping from features to the target label. Question 2. In the context of AI, what does the term “transformer” primarily refer to? A) A hardware accelerator for neural networks B) A type of recurrent neural network architecture C) A deep learning architecture that relies on self-attention mechanisms D) A data preprocessing technique for scaling features Answer: C Explanation: Transformers use self-attention to process sequences in parallel, forming the basis of many large language models. Question 3. Which of the following is NOT a typical component of the Machine Learning lifecycle? A) Data ingestion B) Model training C) Model decommissioning D) Model monitoring Answer: C Explanation: While models may eventually be retired, “model decommissioning” is not a standard lifecycle stage; the core stages are ingestion, training, evaluation, deployment, and monitoring. Question 4. In feature engineering, which technique is best suited for converting a categorical variable with many levels into a numeric representation while preserving ordinal relationships?
A) One-hot encoding B) Label encoding C) Frequency encoding D) Binary encoding Answer: B Explanation: Label encoding assigns an integer to each category, preserving order if the categories have an inherent ranking. Question 5. Which metric is most appropriate for evaluating a binary classifier when the classes are highly imbalanced? A) Accuracy B) Precision C) Recall D) F1-Score Answer: D Explanation: The F1-Score balances precision and recall, providing a better assessment than accuracy when class distribution is skewed. Question 6. A model that performs well on training data but poorly on unseen data is likely suffering from: A) Underfitting B) Overfitting C) High bias D) Data leakage Answer: B Explanation: Overfitting occurs when a model captures noise in the training set, leading to poor generalization. Question 7. Which MLOps practice helps ensure that a newly deployed model does not degrade user experience? A) A/B testing B) Data versioning C) Hyperparameter sweep
Question 11. Which of the following networking components isolates traffic within a cloud VPC at the subnet level? A) Security Group B) Network ACL (NACL) C) Route Table D) Internet Gateway Answer: B Explanation: NACLs operate at the subnet level, controlling inbound and outbound traffic for all resources within that subnet. Question 12. In Azure, which role-based access control (RBAC) element defines a set of permissions that can be assigned to users, groups, or service principals? A) Role Definition B) Role Assignment C) Policy D) Management Group Answer: A Explanation: A Role Definition (or role) specifies the actions that are allowed; it is then assigned to a principal via a Role Assignment. Question 13. Which cloud cost-optimization technique involves purchasing capacity for a specific instance type over a one- or three-year term? A) Spot Instances B) Savings Plans C) Reserved Instances D) Elastic Scaling Answer: C Explanation: Reserved Instances lock in a discounted rate for a chosen instance type for a long-term commitment. Question 14. Which of the following cryptographic algorithms is asymmetric? A) AES
Answer: C Explanation: RSA uses a public-private key pair, making it an asymmetric encryption algorithm. Question 15. The principle of “least privilege” is best implemented through which IAM feature? A) Multi-factor authentication (MFA) B) Role-based access control (RBAC) C) Single sign-on (SSO) D) Password policies Answer: B Explanation: RBAC assigns only the permissions necessary for a role, embodying the least-privilege principle. Question 16. Which security model assumes that no user or device is trusted by default, even if inside the network perimeter? A) Defense-in-Depth B) Zero Trust C) Bell-LaPadula D) Clark-Wright Answer: B Explanation: Zero Trust requires verification for every access request, regardless of location. Question 17. A hash function is primarily used for: A) Encrypting data for confidentiality B) Ensuring data integrity and creating digital fingerprints C) Generating public-private key pairs D) Compressing large files Answer: B
A. Transactional logs B. Time-series sensor data C. Structured sales records D. Unstructured video files Answer: C Explanation: Columnar warehouses excel at aggregating structured, relational data such as sales records. Question 22. In SQL, which clause is used to filter rows after grouping has been performed? A) WHERE B) HAVING C) ORDER BY D) LIMIT Answer: B Explanation: HAVING applies conditions to aggregated groups, whereas WHERE filters rows before grouping. Question 23. Which Python library is most commonly used for data manipulation and analysis in a data-science workflow? A) NumPy B) Pandas C) Matplotlib D) Scikit-learn Answer: B Explanation: Pandas provides DataFrame structures and functions for cleaning, transforming, and analyzing tabular data. Question 24. When performing dimensionality reduction, which technique projects data onto orthogonal axes that capture maximum variance? A) t-SNE B) PCA C) LDA
D) Autoencoder Answer: B Explanation: Principal Component Analysis (PCA) creates orthogonal components ordered by explained variance. Question 25. In time-series forecasting, which method explicitly models seasonality and trend components? A) ARIMA B) Exponential Smoothing (ETS) C) Linear Regression D) K-Nearest Neighbors Answer: B Explanation: ETS (Error, Trend, Seasonal) models separate components, handling both trend and seasonality. Question 26. Which chart type is most appropriate for showing the distribution of a single continuous variable? A) Bar chart B) Pie chart C) Histogram D) Scatter plot Answer: C Explanation: Histograms display frequency counts across intervals, revealing the shape of a continuous distribution. Question 27. In Tableau, which feature allows you to combine multiple data sources into a single view without blending? A) Data Extract B) Data Union C) Data Join D) Data Scaffold Answer: C Explanation: Joins merge tables on common fields, enabling unified analysis across sources.
B) git checkout -b C) git merge D) git pull Answer: B Explanation: git checkout -b both creates the branch and checks it out. Question 32. Which CI/CD tool is native to GitHub and enables workflow automation via YAML files? A) Jenkins B) GitLab CI C) GitHub Actions D) Azure Pipelines Answer: C Explanation: GitHub Actions uses repository-level YAML files to define CI/CD pipelines. Question 33. What is the primary purpose of a Dockerfile’s EXPOSE instruction? A) Set environment variables inside the container B) Define the default command to run C) Document which ports the container listens on D) Copy files from the host into the image Answer: C Explanation: EXPOSE informs users and orchestration tools about the container’s listening ports. Question 34. In Kubernetes, which object ensures that a specified number of pod replicas are always running? A) Service B) Deployment C) ConfigMap D) PersistentVolume Answer: B
Explanation: Deployments manage replica sets, automatically maintaining the desired number of pod replicas. Question 35. Which SRE metric quantifies the acceptable level of service unavailability? A) Service Level Indicator (SLI) B) Service Level Objective (SLO) C) Error Budget D) Mean Time Between Failures (MTBF) Answer: B Explanation: An SLO defines the target performance level (e.g., 99.9% uptime) that the service must meet. Question 36. Which of the following is a key benefit of using Infrastructure as Code (IaC)? A) Manual configuration of servers for flexibility B) Faster provisioning through automated scripts C) Reduced need for version control D) Elimination of security testing Answer: B Explanation: IaC automates resource creation, enabling rapid, repeatable provisioning. Question 37. In reinforcement learning, what does the “reward signal” represent? A) The loss function to be minimized B) The gradient used for backpropagation C) Feedback indicating how good an action was D) The state transition probability matrix Answer: C Explanation: The reward signal quantifies the immediate benefit of an action, guiding the agent’s policy learning.
C) Standardization (z-score scaling) D) Data augmentation Answer: C Explanation: K-Means relies on Euclidean distance; standardizing features prevents those with larger ranges from dominating the distance metric. Question 42. In a supervised classification problem, which loss function is most appropriate for multi-class problems? A) Mean Squared Error B) Hinge Loss C) Cross-Entropy Loss D) Absolute Error Loss Answer: C Explanation: Cross-entropy (softmax) loss measures the divergence between predicted probabilities and true class labels for multi-class tasks. Question 43. Which of the following describes “concept drift” in deployed ML models? A) Model parameters change due to hardware degradation B) Training data becomes larger over time C) The statistical properties of input data change after deployment D) The model’s source code is refactored Answer: C Explanation: Concept drift occurs when the relationship between features and target evolves, causing model performance to degrade. Question 44. Which cloud service provides a fully managed serverless relational database that automatically scales compute and storage? A) Amazon RDS B) Azure SQL Database (serverless tier) C) Google Cloud SQL D) Amazon Aurora Serverless Answer: D
Explanation: Aurora Serverless automatically adjusts capacity based on workload, offering a serverless relational database. Question 45. In AWS IAM, which entity can assume a role to gain its permissions? A) User only B) Group only C) Service principal or another role D) Policy document Answer: C Explanation: Roles are assumed by IAM users, other roles, or AWS services (service principals) to inherit the role’s permissions. Question 46. Which encryption method is typically used to protect data at rest in cloud object storage? A) TLS B) HTTPS C) Server-Side Encryption (SSE) D) VPN tunneling Answer: C Explanation: SSE encrypts objects when stored, ensuring data at rest is protected. Question 47. Which compliance framework focuses specifically on protecting health-information in the United States? A) GDPR B) PCI-DSS C) HIPAA D) ISO 27001 Answer: C Explanation: HIPAA sets standards for safeguarding Protected Health Information (PHI) in the U.S.
C) Encrypting tables for security D) Partitioning data across multiple servers Answer: B Explanation: Normalization structures tables to eliminate redundancy and maintain consistency. Question 52. Which of the following is a visual representation best suited for showing hierarchical relationships? A) Scatter plot B) Pie chart C) Tree map D) Box plot Answer: C Explanation: Tree maps display hierarchical data using nested rectangles, conveying size and structure. Question 53. Which Agile ceremony is primarily responsible for planning the work for the next sprint? A) Daily Stand-up B) Sprint Review C) Sprint Retrospective D) Sprint Planning Answer: D Explanation: Sprint Planning defines the sprint goal and selects backlog items for the upcoming iteration. Question 54. In Terraform, which file defines the desired state of cloud resources? A) variables.tf B) outputs.tf C) main.tf D) terraform.lock.hcl Answer: C
Explanation: main.tf (or any .tf file) contains resource blocks describing the intended infrastructure state. Question 55. Which CI tool uses a “pipeline as code” approach with a Jenkinsfile written in Groovy? A) Travis CI B) CircleCI C) Jenkins D) Bamboo Answer: C Explanation: Jenkins pipelines are defined in a Jenkinsfile using Groovy DSL. Question 56. What is the primary purpose of a service mesh in a microservices architecture? A) Store persistent data for services B) Manage inter-service communication, security, and observability C) Deploy containers onto virtual machines D) Automate code compilation Answer: B Explanation: Service meshes (e.g., Istio) provide a dedicated layer for traffic management, mutual TLS, and telemetry. Question 57. Which of the following best describes canary deployment? A) Deploying a new version to 100% of users instantly B) Rolling back to the previous version after a failure C) Releasing a new version to a small subset of users before full rollout D) Deploying multiple versions simultaneously for A/B testing Answer: C Explanation: Canary deployments expose a small portion of traffic to the new version, allowing validation before wider release. Question 58. In the context of AI ethics, what does model interpretability aim to provide?
D) Lightsail Answer: C Explanation: Fargate abstracts server management, allowing you to run containers directly. Question 62. Which Azure feature enables role-based access control at the resource-group level? A) Azure Policy B) Management Groups C) Role Assignments D) Network Security Groups Answer: C Explanation: Role assignments can be scoped to subscriptions, resource groups, or individual resources, enforcing RBAC. Question 63. In GCP, which storage class is optimized for frequently accessed data with low latency? A) Nearline B) Coldline C) Standard D) Archive Answer: C Explanation: Standard storage offers high availability and low latency for hot data. Question 64. Which of the following is a stateless authentication mechanism commonly used for APIs? A) OAuth 2.0 Authorization Code Flow B) JWT (JSON Web Token) C) Basic Auth over HTTPS D) Session-based cookies Answer: B Explanation: JWTs contain all necessary claims and are not stored server-side, making them stateless.
Question 65. What does CI stand for in the DevOps pipeline? A) Continuous Inspection B) Continuous Integration C) Code Isolation D) Cloud Initiation Answer: B Explanation: CI refers to the practice of frequently merging code changes into a shared repository. Question 66. In Kubernetes, which object abstracts a set of Pods and provides a stable IP address and DNS name? A) ReplicaSet B) Service C) Ingress D) ConfigMap Answer: B Explanation: A Service defines a logical set of Pods and a policy by which to access them, offering a stable endpoint. Question 67. Which monitoring metric would most directly indicate CPU saturation on a server? A) Disk I/O latency B) Network throughput C) CPU utilization percentage near 100% D) Memory free percentage Answer: C Explanation: High CPU utilization approaching 100% signals that the CPU is saturated. Question 68. In the context of data pipelines, what does ELT stand for? A) Extract, Load, Transform B) Encode, Load, Transfer