Quantum kernel methods, variational circuits, and volatility modelling: a reference assessment of the current state of the field.
ARIMA, GARCH, LSTM, and the limits of conventional time-series analysis.
Mapping financial data into quantum feature spaces. Key areas include: Quantum feature maps and quantum support vector machines for classification tasks; Encoding financial time series into quantum Hilbert spaces for regime analysis; Detecting regime changes in non-stationary market data using quantum kernel evaluations.
Parameterised circuits as an alternative to deep learning regression. Key areas include: Parameterised quantum circuits for regression on financial time-series data; Quantum reservoir computing approaches (Fujii and Nakajima 2017) and their applicability; Comparison with classical neural approaches: where variational circuits may generalise better on small training sets.
Encoding, training, and benchmarking against classical baselines. Key areas include: Encoding historical price and volume data into quantum circuit inputs; Training a variational circuit on market data and tuning ansatz depth; Measuring prediction accuracy against an LSTM baseline on the same dataset.
Stochastic volatility, implied surfaces, and calibration. Key areas include: Quantum approaches to stochastic volatility modelling and Monte Carlo acceleration; Implied volatility surface fitting using quantum kernel ridge regression; Quantum-enhanced calibration of pricing models: current results and constraints.
What works, what does not, and the barren plateau problem. Key areas include: Published benchmarks: where QML matches, beats, or loses to classical ML today; The barren plateau problem (McClean et al. 2018) and its implications for circuit scalability; Structuring a research pilot with realistic expectations for 2026 to 2028.
Q&A and Pilot Planning: this session covers the core principles and technical underpinnings relevant to the subject area.
Discuss this topic with senior peers.
