AIF-C01 Question 176
Select 3A company is building a machine learning model to predict customer preferences for an e-commerce platform. The team wants to ensure that the dataset used for training the model is inclusive, diverse, and balanced. Which of the following actions should the team take to achieve this goal?
- A
Include data representing all customer demographics, such as age, gender, and location.
- B
Remove data points from underrepresented groups to maintain simplicity in the dataset.
- C
Collect data from curated and reliable sources to ensure quality and accuracy.
- D
Ensure that the dataset contains equal representation of all classes or categories in the target variable.
- E
Focus only on collecting data from the company's most profitable customer segments.
Show answer and explanation
Correct answers: A, C, D
Explanation
To build an inclusive, diverse, and balanced dataset, it is essential to represent all customer demographics, use reliable and curated data sources, and ensure equal representation of all categories in the target variable. These practices help reduce bias, improve model fairness, and ensure that predictions are applicable to a wide range of customers.
- A. Correct.
Including data representing all customer demographics ensures diversity and makes the model more inclusive, reducing the risk of bias.
- B. Incorrect.
Removing data points from underrepresented groups reduces inclusivity and diversity, leading to biased model predictions. This is not a recommended action.
- C. Correct.
Using curated and reliable data sources improves data quality and ensures that the dataset is representative and accurate.
- D. Correct.
Ensuring equal representation of all classes or categories in the target variable creates a balanced dataset, reducing bias and improving model performance.
- E. Incorrect.
Focusing only on the company's most profitable customer segments narrows the dataset and reduces diversity, leading to a biased model.