MLA-C01 Question 123
Single answerAn e-commerce company wants to predict the likelihood of customers purchasing a product after adding it to their cart. The data consists of structured numerical features such as time spent on the website, number of items in the cart, and customer demographics. You have limited labeled data and need a model that generalizes well to unseen data. Which modeling approach is the most appropriate in this scenario?
- A
Linear regression
- B
Logistic regression
- C
Random forest
- D
K-means clustering
Show answer and explanation
Correct answer: B
Explanation
Logistic regression is well-suited for binary classification problems, especially when the dataset is small and consists of structured numerical data. It is efficient, interpretable, and less prone to overfitting compared to more complex models like random forests. This makes it the most appropriate modeling approach for predicting the likelihood of a purchase based on the given features.
- A. Incorrect.
Linear regression is used for regression tasks to predict continuous numerical values. Since the problem involves predicting the likelihood (a probability) of a binary outcome (purchase or not), linear regression is not a suitable choice.
- B. Correct.
Logistic regression is ideal for binary classification tasks where the output is a probability between 0 and 1. It also works well with structured numerical data and is less prone to overfitting in scenarios with limited labeled data.
- C. Incorrect.
Random forest is a powerful ensemble method, but it may be prone to overfitting when the dataset is small. While it could work, logistic regression is more efficient and interpretable for this problem.
- D. Incorrect.
K-means clustering is a clustering algorithm used for unsupervised learning tasks. Since this is a supervised classification problem with labeled data, K-means is not suitable.