AWS AI Practitioner
A company is training an ML model using a dataset that primarily contains data from North American customers. The company wants to ensure the model works fairly for customers from all regions globally. Which step should the company take to address this?
A
Balance class representation in the dataset by including data from diverse global customer regions
✓ Correcta
B
Remove all regional features from the dataset to make the model region-agnostic
C
Use a pre-trained model that was trained on North American data
D
Reduce the size of the training dataset to minimize regional bias
Explicación
Balancing class representation by including data from diverse global customer regions ensures the model has sufficient training examples from all regions it will serve. This reduces geographic bias and improves model fairness and accuracy for customers worldwide.