Google Professional Machine Learning Engineer Question 79
Select 3Google Cloud PlatformYou are a machine learning engineer tasked with building a model to predict customer churn for a subscription-based service. Your dataset contains tabular data with various features such as customer demographics, subscription details, and usage patterns. You decide to use Google Cloud AutoML Tables to train the model. Which of the following steps are necessary to ensure the best results when using AutoML Tables?
- A
Ensure all categorical data is one-hot encoded before uploading the dataset to AutoML Tables.
- B
Split the dataset into training, validation, and test sets before uploading to AutoML Tables.
- C
Provide a clear objective column (label) indicating the target variable for prediction.
- D
Enable feature explainability to better understand the importance of features.
- E
Verify that the dataset is uploaded in a supported format, such as CSV or BigQuery table.
Show answer and explanation
Correct answers: C, D, E
Explanation
To achieve the best results with AutoML Tables, it is important to provide a clear objective column as the target variable, enable feature explainability for interpretability, and ensure the dataset is in a supported format. AutoML Tables handles preprocessing tasks like one-hot encoding and dataset splitting automatically, so these steps are not required to be performed manually.
- A. Incorrect.
Incorrect: AutoML Tables automatically handles preprocessing tasks such as one-hot encoding for categorical data. You do not need to manually perform this step.
- B. Incorrect.
Incorrect: AutoML Tables automatically handles dataset splitting into training, validation, and test sets. Manually performing this step is not required.
- C. Correct.
Correct: AutoML Tables requires a clear objective column (label) to train a supervised learning model. Without this, the model cannot learn the target prediction task.
- D. Correct.
Correct: Enabling feature explainability in AutoML Tables helps provide insights into which features are most influential in the predictions, improving interpretability.
- E. Correct.
Correct: AutoML Tables requires the dataset to be uploaded in a supported format, such as a CSV file or a BigQuery table, to process the data correctly.