A strategy employed in machine learning to enhance the size and quality of training datasets, thereby improving the performance and generalizability of models. It involves creating modified copies of existing data or generating new data points.

• Image recognition
• Self-driving cars
• Anomaly detection
• Healthcare analytics
• Voice recognition
• Natural Language Processing (NLP)
• Weather prediction

For a computer vision model trained on images of dogs, data augmentation could involve transforming the images to simulate different lighting conditions, angles, or breeds.

• Defining the data vision and strategy