Google Professional Machine Learning Engineer Question 1
Single answerGoogle Cloud PlatformYou are a machine learning engineer tasked with creating a predictive model for customer churn using BigQuery ML. The dataset is stored in BigQuery, and you need to include categorical features like 'subscription type' and 'customer region' in your model training. What should you do to ensure these features are properly handled by BigQuery ML?
- A
Manually encode the categorical features into numeric values before training the model.
- B
Use BigQuery ML's automatic feature preprocessing to handle categorical features.
- C
Exclude categorical features from the training dataset as BigQuery ML does not support them.
- D
Train the model first and then preprocess categorical features at the prediction stage.
Show answer and explanation
Correct answer: B
Explanation
BigQuery ML is designed to handle categorical features automatically during model training, eliminating the need for manual encoding or exclusion of such features. By enabling automatic feature preprocessing, BigQuery ML simplifies the workflow for machine learning engineers, ensuring that categorical features are processed appropriately without additional effort.
- A. Incorrect.
This is incorrect because BigQuery ML automatically handles categorical features through feature preprocessing, so manual encoding is unnecessary.
- B. Correct.
This is correct because BigQuery ML provides automatic feature preprocessing, including handling categorical features through techniques like one-hot encoding or embeddings.
- C. Incorrect.
This is incorrect because BigQuery ML supports categorical features and can process them automatically during training.
- D. Incorrect.
This is incorrect because preprocessing categorical features after training would result in inconsistent inputs for the model, leading to incorrect predictions.