• R.Deepa
• Abhay Pramod
• Benil R
• Sidharth M
• Harish Charan U

In our proposed work, we present a novel approach to address the increasing demand for large, high-quality datasets in machine learning (ML), particularly in the healthcare domain. Synthetic data is generated using four methods, including Copula and Generative Adversarial Networks (GANs), and evaluated for its applicability in lung cancer risk factor analysis. Incremental Ensemble Learning models, comprising Adaptive Random Forest classifiers, Softmax Regressor, and K-Nearest Neighbors (KNN), are employed to assess classification performance using synthetic versus real data. Pearson's correlation coefficient is utilized to measure data similarity, revealing a strong relationship between higher correlation and improved model performance. Among the methods, GAN-generated data demonstrated superior performance and was most challenging to distinguish from real data. Furthermore, the concept is extended to classify medical imaging datasets as real or synthetic. For real datasets, a watermark is embedded into patient reports generated from healthcare data to ensure authenticity and security. Synthetic datasets are excluded from watermarking to support research and simulation purposes. This integration of synthetic data classification, patient report generation, and watermarking enhances data reliability, promotes secure healthcare practices, and facilitates advancements in ML- based medical applications.

Synthetic data, Machine learning, Healthcare, Generative Adversarial Networks (GANs), Incremental Ensemble Learning, Data similarity, Watermarking, Data reliability.

"DIFFERENTIATION AND TAGGING OF REAL Vs SYNTHETIC DATA", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 4, page no.l528-l536, April-2025, Available :http://www.jetir.org/papers/JETIR2504B66.pdf

"DIFFERENTIATION AND TAGGING OF REAL Vs SYNTHETIC DATA", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 4, page no. ppl528-l536, April-2025, Available at : http://www.jetir.org/papers/JETIR2504B66.pdf