icSmartGrid 2025 — Clustering for grid stability prediction

Ümit Şentürk presented our work on unsupervised clustering for stability prediction in smart grid systems, and Batuhan Hangın presented a classical–quantum transfer learning model for disturbance detection in power systems.

The transfer-learning result is the one I keep thinking about: pre-training the classical encoder on abundant simulated data, then fine-tuning only the variational circuit on scarce real disturbance events, recovers most of the benefit of a fully quantum pipeline at a fraction of the circuit depth. On near-term hardware, circuit depth is the budget.