GK Questions

Post-incident review captures lessons learned: what happened, response effectiveness, root causes, improvements. Critical for continuous security program maturity.

Email security combines: SPF/DKIM for authentication, ML for content analysis, heuristics for known patterns. Layered defense critical for phishing protection.

A/B testing randomly assigns users to model variants, measuring metrics like conversion, engagement. Critical for data-driven model iteration and business value validation.

MLOps tools: DVC versions data and models, MLflow tracks experiments and deployments. Combined with Git for end-to-end reproducibility. Critical for enterprise ML governance.

Reinforcement Learning from Human Feedback (RLHF) trains reward models from human preferences, then optimizes LLM to maximize rewards. Critical for aligning AI with human values.

Prompt engineering: clear instructions, examples, formatting, chain-of-thought. Low-cost method to improve LLM outputs without retraining. Critical for practical LLM applications.

Structured output: function calling defines schemas for LLM responses, parsing libraries validate/extract structured data. Critical for integrating LLMs with APIs and databases.

WTO e-commerce negotiations and OECD digital economy work address cross-border data flows, digital taxation, consumer protection. India actively participates in shaping global digital trade rules.

Effective tech policy: proportionate to actual risks, technology-neutral (focus on outcomes), risk-based (higher scrutiny for higher-risk activities). Critical for balanced policy design.

PLI for IT hardware (2021) provides incentives on incremental sales of laptops, tablets, servers manufactured in India. Aims to reduce imports, create jobs, strengthen electronics ecosystem.

CERT-In handles cyber incident response; NCIIPC (National Critical Information Infrastructure Protection Centre) protects critical infrastructure. Both under MeitY for cybersecurity governance.

Digital India encompasses UPI, Aadhaar, DigiLocker for inclusive digital services. UPI enables instant, interoperable payments driving financial inclusion. Critical for digital economy questions.

Imbalance handling: oversampling (SMOTE) increases minority, undersampling reduces majority, class weighting adjusts loss function. Choice depends on data size and model type.

Data leakage: using future data, target information, or test set in training/preprocessing. Causes overfitting and poor generalization. Critical to prevent via proper data splitting and pipeline design.

Regression metrics: RMSE (Root Mean Square Error) penalizes large errors, MAE for robustness, R² for variance explained. Choice depends on business impact of errors.

Categorical encoding: label encoding for ordinal variables, one-hot for nominal, target encoding for high-cardinality. Choice impacts model performance and interpretability.

Stateless microservices store session data externally (Redis, database), enabling horizontal scaling by adding instances. Critical for elastic, resilient cloud applications.

Multi-region architecture distributes workload across geographic regions, enabling failover during regional disasters. Critical for business continuity and global user experience.

Serverless compute (AWS Lambda, GCP Cloud Functions) executes code in response to events: API calls, file uploads, messages. Auto-scaling, pay-per-execution pricing. Critical for event-driven architectures.

Cloud Security Alliance Cloud Controls Matrix provides comprehensive security controls for cloud services, mapped to standards like ISO 27001, NIST. Critical for cloud security governance.