Mehran Pashaei

I’m a Machine Learning Engineer working on deep learning and signal processing for multivariate time-series data in power systems. At FAU Erlangen–Nürnberg’s Pattern Recognition Lab, I develop methods for fault detection, fault classification, and fault localization using high-frequency voltage and current signals.

I build and maintain preprocessing and training pipelines for multivariate time-series (segmentation, normalization, feature extraction, dataset preparation), then train and benchmark models on simulated and real-world grid data with a focus on robustness and generalization. I also label partially unlabeled disturbance datasets using signal-processing and protection-domain indicators, and stress-test models under missing measurements, reduced sampling, and communication loss.