Google Professional Machine Learning Engineer Question 62
Select 2Google Cloud PlatformYou are tasked with building a machine learning model using Google Cloud AutoML. Your dataset consists of customer transaction data with categorical and numerical features. To ensure optimal performance and compatibility with AutoML, which steps should you take during data preparation?
- A
Ensure all categorical features are represented as numerical values, such as one-hot encoding or integer encoding.
- B
Remove all missing values from the dataset to avoid errors during AutoML training.
- C
Label your dataset with the target variable for supervised learning tasks.
- D
Optimize the dataset by manually selecting the most relevant features using statistical methods.
- E
Split the dataset into training, validation, and test sets before uploading it to AutoML.
Show answer and explanation
Correct answers: A, C
Explanation
To prepare data for AutoML, you must ensure categorical features are encoded numerically and that the dataset is labeled for supervised learning tasks. AutoML handles missing values, feature selection, and dataset splitting, simplifying the data preparation process.
- A. Correct.
Correct: AutoML requires categorical features to be converted into numerical representations to process the data effectively.
- B. Incorrect.
Incorrect: AutoML can handle missing values by applying its own preprocessing techniques; removing all missing values is not mandatory.
- C. Correct.
Correct: Labeling the dataset is essential for supervised learning tasks, as AutoML needs to know the target variable to train the model.
- D. Incorrect.
Incorrect: AutoML performs automated feature selection and engineering, so manual optimization of features is not necessary.
- E. Incorrect.
Incorrect: AutoML automatically splits the dataset into training, validation, and test sets, so manual splitting is not required.