Which practice ensures ML models are continuously monitored in production?
- CI/CD
- Model Drift Detection
- A/B Testing
- All of these
Answer: All of these
MLOps combines CI/CD, drift detection, A/B testing for production ML.
Filter by category, type, and difficulty. Reading is open for everyone.
Answer: All of these
MLOps combines CI/CD, drift detection, A/B testing for production ML.
Answer: Function
Function calling enables LLMs to invoke APIs, run code, query databases.
Answer: True
PEFT methods update small subset of parameters, reducing compute/memory.
Answer: RAG
Retrieval-Augmented Generation combines LLMs with external knowledge retrieval.
Answer: Thought
Chain-of-Thought prompting encourages LLMs to generate reasoning steps.
Answer: Clustering
Clustering discovers natural groupings in unlabeled data.
Answer: True
Feature engineering: domain knowledge, transformations, interactions, encoding.
Answer: Time Series Forecasting
Time series forecasting models temporal patterns for demand prediction, anomaly detection.
Answer: Serverless
Serverless query engines enable SQL queries on object storage using schema-on-read.
Answer: True
Data lakes store raw structured/unstructured data; warehouses store curated data.
Answer: Parquet
Parquet is columnar storage format optimized for analytics: compression, predicate pushdown.
Answer: Stream
Stream processing handles continuous data flows for real-time insights.
Answer: True
CDC captures database operations and streams to targets for real-time analytics.
Answer: True
IOCs are forensic artifacts indicating potential breaches. Used for proactive blocking.
Answer: True
Micro-segmentation creates granular security zones, containing breaches and limiting attacker movement.
Answer: Both A and B
MFA requires multiple authentication factors; adaptive authentication adjusts requirements based on risk context.
Answer: FAME
FAME provides demand incentives for EVs and charging infrastructure.
Answer: Model / Pipeline
Model parallelism splits model layers across devices; pipeline parallelism stages computation. Critical for scaling LLM training.
Answer: True
Positional encodings add sequence order information to token embeddings, critical for language understanding tasks.
Answer: Attention Mechanism
Self-attention computes weighted relationships between tokens, enabling context-aware processing. Foundation of Transformer success.