Google Professional Machine Learning Engineer exam dumps

Google Professional Machine Learning Engineer practice question 136 of 522

Professional Machine Learning Engineer. Professional level, Google Cloud. Free question with the correct answer and a full explanation.

Google Professional Machine Learning Engineer Question 136

Select 2Google Cloud Platform

You are tasked with deploying a machine learning model on Vertex AI to classify customer feedback text documents. The input data consists of large text files stored in Google Cloud Storage (GCS). You want to preprocess the data and perform batch predictions using the model deployed on Vertex AI. Which of the following steps should you take to accomplish this?

  1. A

    Use a Dataflow pipeline to preprocess the text data and save the results back to GCS in a format supported by Vertex AI Batch Prediction.

  2. B

    Deploy the model to a Vertex AI Prediction Endpoint and use it for real-time predictions on the text documents.

  3. C

    Use the Vertex AI Batch Prediction service to process the preprocessed data stored in GCS.

  4. D

    Directly load the raw text files from GCS into Vertex AI Batch Prediction without preprocessing.

  5. E

    Store the preprocessed data in BigQuery and use the BigQuery ML batch prediction feature.

Show answer and explanation

Correct answers: A, C

Explanation

To perform batch predictions in Vertex AI, the input data must be preprocessed into a supported format like JSONL. A Dataflow pipeline can be used for this preprocessing step, and the resulting data should be stored in GCS. Vertex AI Batch Prediction can then process this data efficiently. Real-time prediction endpoints and BigQuery ML are not appropriate for this batch inference scenario.

  • A. Correct.

    Correct. Preprocessing the text data using a Dataflow pipeline ensures that the data is transformed into a format compatible with Vertex AI Batch Prediction, such as JSONL.

  • B. Incorrect.

    Incorrect. Real-time predictions using a Vertex AI Prediction Endpoint are useful for single or small-scale predictions but not for batch processing of large datasets.

  • C. Correct.

    Correct. Vertex AI Batch Prediction is designed for large-scale inference on preprocessed data stored in GCS, making it the appropriate choice for this scenario.

  • D. Incorrect.

    Incorrect. Raw text files must be preprocessed into a format supported by Vertex AI Batch Prediction, such as JSONL, before they can be used.

  • E. Incorrect.

    Incorrect. While BigQuery ML supports batch predictions, this is not relevant to the scenario as the model is deployed on Vertex AI, not BigQuery ML.

Timed practice exam

Take a Google Professional Machine Learning Engineer practice test under exam conditions

60 questions in 120 minutes, drawn from this bank, with a score report and a per-question review when you finish.

Start timed exam