MLA-C01 Question 118
Select 4You are developing a machine learning model to predict customer churn for a subscription-based service. The dataset contains numerical, categorical, and textual features. You decide to use Amazon SageMaker for the model training and deployment. Which steps should you perform to ensure the dataset is ready for training?
- A
Convert categorical features into numerical representations by using techniques like one-hot encoding or embedding.
- B
Normalize numerical features to ensure they are on the same scale.
- C
Remove all textual features as they cannot be processed by machine learning models.
- D
Partition the dataset into training, validation, and testing subsets.
- E
Replace missing values in the dataset by applying appropriate imputation strategies.
Show answer and explanation
Correct answers: A, B, D, E
Explanation
To prepare a dataset for training a machine learning model, you must preprocess the data appropriately. This includes converting categorical features into numerical formats, normalizing numerical features, partitioning the dataset, and addressing missing values. Removing textual features is unnecessary, as there are techniques to process them, such as tokenization or embedding. Proper preprocessing ensures the model can learn efficiently and perform well on unseen data.
- A. Correct.
Correct. Categorical features need to be converted into numerical representations, as most machine learning algorithms cannot handle categorical data directly.
- B. Correct.
Correct. Normalizing numerical features ensures they are on the same scale, which can improve the performance of many machine learning models.
- C. Incorrect.
Incorrect. Textual features can be processed using techniques such as tokenization or embedding, and removing them outright may result in loss of valuable information.
- D. Correct.
Correct. Partitioning the dataset into training, validation, and testing subsets is essential for building a robust machine learning model.
- E. Correct.
Correct. Handling missing values is a critical step in preprocessing to ensure the model can effectively learn from the data.