ARTIFICIAL DATA GENERATION USING GANS FOR PRIVACY-PRESERVING MACHINE LEARNING

The fast growth of data-driven applications has increased the need for large-scale datasets to train
machine learning models. However, the use of real-world data often raises serious privacy and
security concerns, especially in sensitive domains such as banking, healthcare, finance, and
education. Synthetic data generation has evolved as an emerging solution to address these
challenges. This paper presents a simple and effective approach for generating synthetic data using
Generative Adversarial Networks (GANs) to enable privacy-preserving machine learning. A
GAN-based framework is implemented to generate synthetic samples that closely resemble real data while ensuring that sensitive information is not directly exposed. Experimental results using a standard benchmark dataset demonstrate that machine learning models trained on synthetic data achieve equivalent performance to those trained on real data. The study emphasizes the potential of GAN-based synthetic data generation as a practical and beginner-friendly approach for
privacy-aware AI systems.