CERTIFIED-MACHINE-LEARNING-PROFESSIONAL · Question #59
CERTIFIED-MACHINE-LEARNING-PROFESSIONAL Question #59: Real Exam Question with Answer & Explanation
The correct answer is B: Streaming. There's an important issue worth flagging here: the marked correct answer of B (Streaming) conflicts with the standard answer found in ML deployment literature. Most authoritative sources - including Chip Huyen's Designing Machine Learning Systems and major ML engineering courses
Question
Which of the following machine learning model deployment paradigms is the most common for machine learning projects?
Options
- AOn-device
- BStreaming
- CReal-time
- DBatch
- ENone of these deployments
Explanation
There's an important issue worth flagging here: the marked correct answer of B (Streaming) conflicts with the standard answer found in ML deployment literature. Most authoritative sources - including Chip Huyen's Designing Machine Learning Systems and major ML engineering courses - identify Batch (D) as the most common deployment paradigm, since the majority of ML use cases (recommendations refreshed nightly, fraud scoring on historical data, churn prediction) don't require predictions faster than a scheduled job can provide.
If the intended correct answer is D (Batch): Batch deployment is the most common because it is the simplest, cheapest, and most operationally mature pattern - models run on a schedule against stored data. Real-time (C) requires always-on infrastructure and low-latency serving, adding complexity used only when predictions must be immediate (e.g., ad ranking). Streaming (B) sits between batch and real-time, processing data as it arrives in a pipeline, but requires stream-processing infrastructure (Kafka, Flink) that most projects don't justify. On-device (A) is specialized for edge/mobile contexts with privacy or latency constraints.
Memory tip: Think "most common = least infrastructure required." Batch wins because a cron job and a database is all you need - no stream brokers, no low-latency APIs, no special hardware.
I'd recommend double-checking your course materials or instructor, as the answer key may contain an error. If your source definitively says B, ask your instructor to clarify their reasoning.
Community Discussion
No community discussion yet for this question.