Quantum-Enhanced Temporal Embeddings via a Hybrid Seq2Seq Architecture

2026-02-12

AI Summary

This work addresses the challenges of noise sensitivity and limited sample size in financial time series by proposing a hybrid quantum-classical sequence-to-sequence autoencoder, termed QLSTM. It uniquely integrates a NISQ-compatible depth-1 variational quantum circuit into the LSTM gating mechanism to refine the geometry of the latent manifold and enhance temporal embedding expressiveness. Leveraging the representations generated by QLSTM, the authors construct graph structures using an RBF kernel and incorporate momentum strategies (RBF-Graph and RBF-DivMom). Evaluated across 14 rolling windows of S&P 500 data from 2022 to 2025, the approach significantly improves trajectory smoothness, state transition clarity, and sectoral clustering stability compared to classical LSTM, yielding downstream portfolios that consistently outperform benchmark methods in risk-adjusted returns.

Technology Category

• Machine Learning: Quantum Machine Learning
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

This work investigates how shallow, NISQ-compatible quantum layers can improve temporal representation learning in real-world sequential data. We develop a QLSTM Seq2Seq autoencoder in which a depth-1 variational quantum circuit is embedded inside each recurrent gate, shaping the geometry of the learned latent manifold. Evaluated on fourteen rolling S and P 500 windows from 2022 to 2025, the quantum-enhanced encoder produces smoother trajectories, clearer regime transitions, and more stable, sector-coherent clusters than a classical LSTM baseline. These geometric properties support the use of a Radial Basis Function (RBF) kernel for downstream portfolio allocation, where both RBF-Graph and RBF-DivMom strategies consistently outperform their classical counterparts in risk-adjusted terms. Analysis across periods shows that compressed manifolds favor concentrated allocation, while dispersed manifolds favor diversification, demonstrating that latent geometry serves as a regime indicator. The results highlight a practical role for shallow hybrid quantum and classical layers in NISQ-era sequence modeling, offering a reproducible pathway for improving temporal embeddings in finance and other data-limited, noise-sensitive domains.

Problem

• quantum-enhanced
• temporal embeddings
• NISQ
• sequence modeling
• representation learning

Innovation

• Quantum-Enhanced Temporal Embeddings
• Hybrid Seq2Seq Architecture
• NISQ-Compatible Quantum Layers
• Latent Manifold Geometry
• QLSTM