Quantum feature maps, QSVM, and non-linear seasonal pattern detection benchmarked against classical RBF kernels.
Where linear decomposition and Fourier methods break down in logistics.
Havlicek et al. (2019) and the quantum kernel advantage hypothesis. Key areas include: Quantum feature maps: encoding classical data into quantum Hilbert space via parameterised circuits; Quantum kernel estimation: computing inner products in exponentially large feature spaces without explicit state preparation; The kernel advantage hypothesis: when quantum kernels provably separate data that classical kernels cannot (Liu et al. 2021).
Applying quantum kernels to non-linear seasonality and regime changes. Key areas include: Multi-scale seasonality: weekly, monthly, and annual cycles with year-over-year drift in logistics demand patterns; Regime change detection: quantum kernels for identifying structural breaks in seasonal behaviour after market shifts; Cross-product seasonal correlation: using QSVM to cluster SKUs by seasonal profile similarity in high-dimensional feature spaces.
Building a quantum kernel model for seasonal logistics data. Key areas include: Constructing a quantum kernel using Qiskit on a 12-qubit simulated backend with ZZ feature maps; Training a QSVM for seasonal pattern classification on a 3-year logistics demand dataset; Comparing classification accuracy against RBF-SVM, random forest, and XGBoost baselines on matched test sets.
Circuit depth, feature dimension limits, and when classical kernels suffice. Key areas include: Feature dimension versus qubit count: current NISQ devices limit practical feature maps to roughly 10-20 input dimensions; Training cost: quantum kernel matrix computation requires O(n^2) circuit evaluations, scaling poorly with dataset size; When classical kernels are sufficient: empirical evidence that most logistics seasonal patterns are separable by RBF or polynomial kernels.
Connecting quantum kernel research to production seasonal models.
Q&A and Action Planning: this session covers the core principles and technical underpinnings relevant to the subject area.
Discuss this topic with senior peers.
