CTO presents 'Forecasting with Quantum Reservoir Computing' at Richmond Fed.

FirstQFM Co-founder and CTO Isaiah Hull visited the Federal Reserve Bank of Richmond for CORE Week and gave a research seminar on "Forecasting with Quantum Reservoir Computing."

The seminar introduced quantum reservoir computing as an alternative to sequential machine learning models, workhorse time-series methods, and variational quantum machine learning algorithms. The presentation framed QRC as a reservoir-based feature generator for forecasting, where a quantum system produces nonlinear features and a classical linear readout performs the final prediction step.

For forecasting use cases, the talk emphasized QRC's near-term practicality. Unlike quantum algorithms that depend on fault-tolerant hardware, QRC is designed for hybrid computation on noisy intermediate-scale quantum devices. It avoids deep variational optimization, keeps training in a classical convex routine, and can be applied to forecasting, classification, and anomaly detection tasks.

The seminar also discussed where QRC fits within public-sector and financial forecasting workflows. Examples included realized volatility forecasting and chaotic time-series prediction, along with design choices around input encoding, measurement strategy, fading memory, and real-time evaluation constraints.