Synthetic Data
Artificially generated data that mimics real-world data patterns, used for training when real data is scarce or sensitive.
Overview
Synthetic data is artificially generated data designed to have the same statistical properties as real-world data. It can be created using rule-based methods, statistical models, or generative AI (GANs, diffusion models, LLMs). Synthetic data addresses data scarcity, privacy concerns, and class imbalance.
Applications
Healthcare uses synthetic patient records for research without privacy risks. Autonomous vehicles use synthetic driving scenarios for rare events. Financial services generate synthetic transactions for fraud detection training. LLM providers use synthetic data for instruction tuning and alignment.