Answer-first summary for fast verification
Answer: It is optimized to minimize the latency of serving predictions., Predictions are returned in the response message.
Online prediction is optimized to minimize the latency of serving predictions and returns predictions directly in the response message. In contrast, batch prediction is designed to handle large volumes of data instances in a single job, supports more complex models, and writes predictions to specified Cloud Storage locations. For more details, refer to [Google Cloud's documentation on prediction types](https://cloud.google.com/ml-engine/docs/prediction-overview#online_prediction_versus_batch_prediction).
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
No comments yet.
What distinguishes online prediction from batch prediction in terms of characteristics?
A
Predictions are written to output files in a Cloud Storage location that you specify.
B
It is optimized to minimize the latency of serving predictions.
C
It is optimized to handle a high volume of data instances in a job and to run more complex models.
D
Predictions are returned in the response message.