Practice Test 3 | Google Cloud Certified Professional Data Engineer | Dumps | Mock Test
Data science team has finally finished implementing a face recognition model to identify visitors entering into the facility building whether they are first-time or frequent visitors. First-time visitors are obligated to register their ID information at the reception to get authorized entrance to the facility’s labs. The model was built using Tensorflow. Head of Data assigned you to deploy the model for production use. Due to the system’s architecture design, the deployed model’s output should be returned in the response message with minimal latency.
Which of the following approaches is best suited in this scenario?
A. Use Cloud Machine Learning Engine to deploy the model. Use batch prediction to pass input data to the model hosted in the cloud.
B. Use Cloud Machine Learning Engine to deploy the model. Use online prediction to pass input data to the model hosted in the cloud.
C. Use Google Kubernetes Engine to deploy the model. Use batch prediction to pass input data to the model hosted in the cloud.
D. Use Google Kubernetes Engine to deploy the model. Use online prediction to pass input data to the model hosted in the cloud.
Correct Answer: B
AI Platform provides two ways to get predictions from trained models: online prediction (sometimes called HTTP prediction), and batch prediction. In both cases, you pass input data to a cloud-hosted machine-learning model and get inferences for each data instance.
Online prediction is optimized for minimal latency. It allows returning the output by a response message.
Options C and D are incorrect. GKE is not a recommended option to deploy the model.
Option A is incorrect: Batch prediction doesn’t support returning the output with a response message. Instead, it writes to Google Storage. Minimal latency is not guaranteed for batch prediction as opposed to online prediction.
Source(s):
Online vs. Batch Prediction:
https://cloud.google.com/ml-engine/docs/tensorflow/online-vs-batch- prediction
Comments are closed, but trackbacks and pingbacks are open.