Saved in:
Bibliographic Details
Main Authors: Hsieh, Tien-Ching, Tsai, Yun-Cheng, Chen, Samuel Yen-Chi
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.11578
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917269994471424
author Hsieh, Tien-Ching
Tsai, Yun-Cheng
Chen, Samuel Yen-Chi
author_facet Hsieh, Tien-Ching
Tsai, Yun-Cheng
Chen, Samuel Yen-Chi
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2602_11578
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Quantum-Enhanced Temporal Embeddings via a Hybrid Seq2Seq Architecture
Hsieh, Tien-Ching
Tsai, Yun-Cheng
Chen, Samuel Yen-Chi
Computational Engineering, Finance, and Science
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.
title Quantum-Enhanced Temporal Embeddings via a Hybrid Seq2Seq Architecture
topic Computational Engineering, Finance, and Science
url https://arxiv.org/abs/2602.11578