Academic outputs and research artifacts.

2026 · BRACIS 2025 · Springer LNCS

Training and Test Machine Learning Models on Encrypted Data: Initial Results and Challenges

Rodrigo Kruger, Jean Paul Barddal, Vinícius Mourão Alves de Souza

This paper studies machine learning over encrypted data using the CKKS homomorphic encryption scheme, evaluating classifier adaptations across medical datasets and analyzing the tradeoffs between privacy, error rates, and processing time.

Kruger, R., Barddal, J. P., & Souza, V. M. A. (2026). Training and Test Machine Learning Models on Encrypted Data: Initial Results and Challenges. In Intelligent Systems (BRACIS 2025), Lecture Notes in Computer Science, vol. 16180. Springer, Cham. https://doi.org/10.1007/978-3-032-15984-7_38

@inproceedings{kruger2026encryptedml,
  title={Training and Test Machine Learning Models on Encrypted Data: Initial Results and Challenges},
  author={Kruger, Rodrigo and Barddal, Jean Paul and Souza, Vinícius Mourão Alves de},
  booktitle={Lecture Notes in Computer Science},
  volume={16180},
  pages={562--577},
  year={2026},
  doi={10.1007/978-3-032-15984-7_38}
}

2024 · 2024 International Joint Conference on Neural Networks (IJCNN)

Peak Prediction in Time Series: Comparing Approaches for Energy High-Load Prediction

Rodrigo Kruger, Abdullah Mueen, Vinícius Mourão Alves de Souza

This conference paper compares approaches for predicting peaks in energy load time series, focusing on high-load prediction as a time-series forecasting problem.

@inproceedings{kruger2024peakprediction,
  title={Peak Prediction in Time Series: Comparing Approaches for Energy High-Load Prediction},
  author={Kruger, Rodrigo and Mueen, Abdullah and Souza, Vinícius Mourão Alves de},
  booktitle={2024 International Joint Conference on Neural Networks (IJCNN)},
  year={2024},
  doi={10.1109/IJCNN60899.2024.10651140}
}