[C5] Hybrid Quantum-Classical Framework for Anomaly Detection in Time Series with QUBO formulation and QAOA
Published in IJCNN, 2025
Recommended citation: M. Casalbore, L. Lavagna, A. Rosato and M. Panella, "Hybrid Quantum-Classical Framework for Anomaly Detection in Time Series with QUBO formulation and QAOA," 2025 International Joint Conference on Neural Networks (IJCNN), Rome, Italy, 2025, pp. 1-8. https://ieeexplore.ieee.org/document/11228152
In this work, we introduce a hybrid quantum-classical framework to address anomaly-detection problems in time-series data using a Quadratic Unconstrained Binary Optimization (QUBO) formulation. The proposed approach integrates density-based and statistical methodologies with the Quantum Approximate Optimization Algorithm (QAOA), providing a versatile framework for anomaly detection across different types of anomalies and applications. Experimental results show competitive accuracy compared with classical techniques while enabling the exploration of alternative solutions through the quantum optimization component. This capability can provide additional insight into the structure of anomalous patterns in time-series data.