MLS-C01 Question 368
Single answerA data scientist is tasked with building a machine learning model to predict customer churn for an e-commerce platform. During the initial analysis, the data scientist finds that the dataset contains categorical, numerical, and missing values. They plan to use Amazon SageMaker to implement the solution. Which machine learning model would be the best choice to handle such data with minimal preprocessing?
- A
Linear Regression
- B
Support Vector Machines (SVM)
- C
Random Forest
- D
Principal Component Analysis (PCA)
Show answer and explanation
Correct answer: C
Explanation
Random Forest is well-suited for datasets with a mix of categorical and numerical features, as well as missing values, because it can handle such complexities without requiring significant preprocessing. This makes it the most appropriate choice for solving the customer churn prediction problem in this scenario.
- A. Incorrect.
Linear Regression is a regression algorithm best suited for predicting continuous numerical outcomes. It is not ideal for handling categorical or missing data directly without significant preprocessing.
- B. Incorrect.
Support Vector Machines (SVM) are effective for classification problems but often require preprocessing to handle categorical features and missing values. This makes it less suitable for the given scenario with minimal preprocessing requirements.
- C. Correct.
Random Forest is a tree-based ensemble learning method that can natively handle both categorical and numerical data, as well as missing values, without extensive preprocessing. This makes it the best choice for the described scenario.
- D. Incorrect.
Principal Component Analysis (PCA) is a dimensionality reduction technique, not a predictive model. It is not suitable for solving a classification problem like customer churn prediction.