In the context of business intelligence, how would you address the challenges posed by imbalanced datasets when developing a machine learning model, and what techniques could be employed to mitigate the impact of skewed class distributions?

Interview question asked to Business Intelligence Engineers interviewing at Amazon Web Services and Amazon: In the context of business intelligence, how would you address the challenges posed by imbalanced datasets when developing a machine learning model, and what techniques could be employed to mitigate the impact of skewed class distributions?.