AI-900 Question 108
Select 3A retail company wants to predict customer churn using machine learning. They have a dataset with customer demographics, purchase history, and customer complaints. Which of the following describes the core concepts required to solve this problem?
- A
Splitting the data into training and testing datasets to evaluate the model
- B
Using labeled data to train a supervised learning model
- C
Deploying the model without testing to avoid delays
- D
Identifying relevant features such as purchase history and customer complaints
- E
Using unsupervised learning to classify customers based on churn likelihood
Show answer and explanation
Correct answers: A, B, D
Explanation
To address the customer churn problem, the company needs to follow core machine learning concepts, such as splitting the data into training and testing datasets for evaluation, using supervised learning for labeled data, and selecting relevant features for the model. These steps ensure the model is trained and evaluated effectively to provide accurate predictions.
- A. Correct.
Splitting the data into training and testing datasets is a fundamental step to evaluate the performance of the machine learning model and ensure it generalizes well to unseen data.
- B. Correct.
Using labeled data to train a supervised learning model is appropriate because predicting customer churn involves a known target variable (churn or no churn).
- C. Incorrect.
Deploying the model without testing would lead to unreliable predictions, as the model's performance would remain unverified.
- D. Correct.
Identifying relevant features, such as purchase history and customer complaints, is critical for training a model that can accurately predict customer churn.
- E. Incorrect.
Unsupervised learning is not suitable in this case, as predicting churn requires labeled data and a target variable, making it a supervised learning problem.