TY  - JOUR
AU  - Mohd, Abdul Kashif
AU  - Saleem, Nasir
AU  - Gogate, Mandar
AU  - Hussain, Adeel
AU  - Dashtipour, Kia
PY  - 2026
DA  - 2026/09/22
TI  - Privacy-Preserving Machine Learning Using Homomorphic Encryption
JO  - ICCK Transactions on Information Security and Cryptography
T2  - ICCK Transactions on Information Security and Cryptography
JF  - ICCK Transactions on Information Security and Cryptography
VL  - 2
IS  - 3
SP  - 159
EP  - 171
DO  - 10.62762/TISC.2026.247194
UR  - https://www.icck.org/article/abs/TISC.2026.247194
KW  - homomorphic encryption
KW  - privacy-preserving machine learning
KW  - CKKS
KW  - logistic regression
KW  - AdaBoost
KW  - encrypted inference
AB  - The exponential growth of machine learning applications in healthcare, finance, and cloud computing has raised significant data privacy concerns, as traditional encryption methods require access to decrypted data for processing, creating security vulnerabilities. Regulatory frameworks including the European Union's General Data Protection Regulation and the Health Insurance Portability and Accountability Act mandate privacy protection throughout the entire data management lifecycle. This research develops a privacy-preserving machine learning framework combining CKKS-based homomorphic encryption with Logistic Regression and AdaBoost classifiers using the UCI Adult Income dataset containing 48,842 records. Arithmetic verification demonstrates CKKS errors of \(2.17 \times 10^{-9}\) for addition, \(4.71 \times 10^{-9}\) for subtraction, and \(7.00 \times 10^{-9}\) for scalar multiplication. Plaintext evaluation shows AdaBoost outperforming Logistic Regression with accuracy of 0.8221, precision of 0.6072, F1 score of 0.6937, and ROC-AUC of 0.9108. Encrypted variants LR+HE and AdaBoost+HE produce results numerically identical to plaintext counterparts across all five evaluation metrics on 6,033 test instances, proving that encryption does not affect predictive performance. The study establishes CKKS-based homomorphic encryption as a secure and effective solution enabling organisations to perform machine learning inference while complying with regulatory requirements.
SN  - 3070-2429
PB  - Institute of Central Computation and Knowledge
LA  - English
ER  - 

@article{Mohd2026PrivacyPre,
  author = {Abdul Kashif Mohd and Nasir Saleem and Mandar Gogate and Adeel Hussain and Kia Dashtipour},
  title = {Privacy-Preserving Machine Learning Using Homomorphic Encryption},
  journal = {ICCK Transactions on Information Security and Cryptography},
  year = {2026},
  volume = {2},
  number = {3},
  pages = {159-171},
  doi = {10.62762/TISC.2026.247194},
  url = {https://www.icck.org/article/abs/TISC.2026.247194},
  abstract = {The exponential growth of machine learning applications in healthcare, finance, and cloud computing has raised significant data privacy concerns, as traditional encryption methods require access to decrypted data for processing, creating security vulnerabilities. Regulatory frameworks including the European Union's General Data Protection Regulation and the Health Insurance Portability and Accountability Act mandate privacy protection throughout the entire data management lifecycle. This research develops a privacy-preserving machine learning framework combining CKKS-based homomorphic encryption with Logistic Regression and AdaBoost classifiers using the UCI Adult Income dataset containing 48,842 records. Arithmetic verification demonstrates CKKS errors of \(2.17 \times 10^{-9}\) for addition, \(4.71 \times 10^{-9}\) for subtraction, and \(7.00 \times 10^{-9}\) for scalar multiplication. Plaintext evaluation shows AdaBoost outperforming Logistic Regression with accuracy of 0.8221, precision of 0.6072, F1 score of 0.6937, and ROC-AUC of 0.9108. Encrypted variants LR+HE and AdaBoost+HE produce results numerically identical to plaintext counterparts across all five evaluation metrics on 6,033 test instances, proving that encryption does not affect predictive performance. The study establishes CKKS-based homomorphic encryption as a secure and effective solution enabling organisations to perform machine learning inference while complying with regulatory requirements.},
  keywords = {homomorphic encryption, privacy-preserving machine learning, CKKS, logistic regression, AdaBoost, encrypted inference},
  issn = {3070-2429},
  publisher = {Institute of Central Computation and Knowledge}
}