Many machine learning projects fail because data scientists don’t have the right data. Techniques such as synthetic data is a novel algorithmic approach to address algorithmic risks. Published at: https://www.eckerson.com/articles/synthetic-data-for-ai-definition-risks-and-strategies