Multivariate Time Series Forecasting with Adaptive Non-Local Observables

2026-07-27

AI Summary

This work addresses the limited expressivity of quantum neural networks in multivariate time series forecasting, which stems from their reliance on fixed local observables. To overcome this limitation, the authors propose the MTSF-ANO hybrid model, which innovatively integrates variational quantum circuits with adaptive non-local observables, substantially enhancing both model expressivity and prediction flexibility. Ablation studies confirm the effectiveness of the non-local observables and circuit design. Evaluated across 20 experimental settings on four ETT datasets, the proposed method achieves top-two performance in 17 cases according to mean squared error (MSE), with improvements of up to 20% over the strongest baseline on the ETTh1 dataset.

Abstract

Multivariate time series forecasting (MTSF) predicts future values of multiple variables from historical data. While quantum neural networks have been increasingly applied to this task, they typically rely on fixed local measurements, which restrict their expressivity. We propose MTSF-ANO, a simple hybrid model for MTSF that integrates variational quantum circuits with adaptive non-local observables (ANO). On the four ETT datasets, MTSF-ANO ranks first or second in MSE in 17 of 20 settings, improving over the strongest baseline by up to 20% on ETTh1, and outperforms or matches its fixed local observable counterpart across all settings. Our ablations show how the quantum circuit design and ANO non-locality affect performance. These results suggest that ANO is a promising direction for quantum time series forecasting.

Problem

• Multivariate Time Series Forecasting
• Quantum Neural Networks
• Local Observables
• Expressivity
• Non-Local Observables

Innovation

• adaptive non-local observables
• variational quantum circuits
• multivariate time series forecasting
• quantum neural networks
• hybrid quantum-classical model