An organization is training a machine learning model to make predictions. What could improve the prediction accuracy of their model?

Increased storage capacity affects how much data can be retained but does not by itself improve the model's predictive accuracy.

Higher network bandwidth speeds up data transfer rates but has no direct effect on the quality or accuracy of a trained ML model.

Machine learning models generalize better when trained on larger and more diverse datasets because they encounter a wider range of input patterns, reducing overfitting and improving predictive performance on unseen data. More training examples allow the model to learn statistical relationships with greater confidence. This is a fundamental principle of supervised learning where data quantity and quality directly drive model accuracy.

Faster CPU processors reduce training time but do not improve the model's accuracy - the same algorithm on more CPUs produces the same quality predictions.