Scholarlink

Electroencephalography (EEG) data is highly sensitive, requiring strong privacy protection in brain–computer interfaces and neurological diagnosis. Homomorphic encryption enables secure EEG classification with deep neural networks, but standard activation functions (e.g., ReLU) are incompatible with encrypted arithmetic, causing severe computational overhead. Polynomial approximations mitigate this issue yet fail to capture EEG’s distinctive spectral patterns, temporal dynamics, and inter-subject variability. We propose a surrogate-assisted genetic algorithm (SA-GA) to optimize activation function approximations for encrypted deep learning models. SA-GA uses fixed-length chromosome encoding and a Gaussian Process surrogate to efficiently search the approximation space while avoiding costly encrypted evaluations. Applied to motor imagery classification (BCI Competition IV Dataset 2a), our method achieves 89.7% accuracy—7.2% higher than conventional polynomial approaches—while remaining feasible for real-time use. This enables practical, privacy-preserving deployment of EEG deep learning systems in clinical and BCI applications.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

SA-GA: Surrogate-Assisted Genetic Algorithm for Optimizing Activation Function Approximations in Privacy-Preserving EEG Classification Networks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Essam Debie,
  • Harshita Harshita

摘要

Electroencephalography (EEG) data is highly sensitive, requiring strong privacy protection in brain–computer interfaces and neurological diagnosis. Homomorphic encryption enables secure EEG classification with deep neural networks, but standard activation functions (e.g., ReLU) are incompatible with encrypted arithmetic, causing severe computational overhead. Polynomial approximations mitigate this issue yet fail to capture EEG’s distinctive spectral patterns, temporal dynamics, and inter-subject variability. We propose a surrogate-assisted genetic algorithm (SA-GA) to optimize activation function approximations for encrypted deep learning models. SA-GA uses fixed-length chromosome encoding and a Gaussian Process surrogate to efficiently search the approximation space while avoiding costly encrypted evaluations. Applied to motor imagery classification (BCI Competition IV Dataset 2a), our method achieves 89.7% accuracy—7.2% higher than conventional polynomial approaches—while remaining feasible for real-time use. This enables practical, privacy-preserving deployment of EEG deep learning systems in clinical and BCI applications.