Institution profile

Wells Fargo, San Francisco, Calif.

Industry researchnorthamerica · us
Official website
Research library13linked papers
Opportunities0open roles
Selected work

Representative Papers

Group Quantization and Mellin Representations of the Heston Model

Jun 11, 2026

This work provides a geometric interpretation of the Heston stochastic volatility model, uncovering the intrinsic mechanism underlying its affine structure. By constructing an augmented local Lie groupoid formulation and introducing, for the first time, group quantization techniques into this framework, the study unifies the pricing operator in coordinate space with the Riccati equations in momentum space. Leveraging the Mellin transform together with geometric representation theory, the approach not only reproduces the classical characteristic function and Riccati solutions but also demonstrates that both arise as dual perspectives of a single geometric construction. This insight endows the Heston pricing formula with a clear and coherent geometric meaning.

0 citationsRead paper

Batched Training for QLSTM vs. QFWP: A System-Oriented Approach to EPC-Aware RMSE-DA

Dec 25, 2025

This study systematically compares quantum long short-term memory (QLSTM) and quantum forgetful wave packet (QFWP) models for daily EUR/USD exchange rate forecasting under equivalent parameter count (EPC) and adjoint-based differentiation. Method: We introduce the first EPC-aligned, numerically reproducible quantum time-series modeling benchmark, incorporating batched tensor parallelism and nonparametric statistical testing (Wilcoxon signed-rank test and Cliff’s delta). Contribution/Results: QFWP consistently outperforms QLSTM across all batch sizes in RMSE and directional accuracy (p ≤ 0.004). QLSTM achieves peak throughput at batch size 64. Forward computation accelerates by 2.2–2.4×; end-to-end training acceleration reaches up to 2×. We uncover asymmetric scalability between forward and backward passes in quantum RNNs and characterize the speed–accuracy Pareto frontier. This work establishes a reproducible benchmark and provides principled model selection guidance for one-dimensional quantum time-series modeling.

0 citationsRead paper

Q-A3C2: Quantum Reinforcement Learning with Time-Series Dynamic Clustering for Adaptive ETF Stock Selection

Dec 25, 2025

To address overfitting from high-dimensional features and the inflexibility of static clustering in dynamic ETF stock selection under evolving market regimes, this paper proposes a quantum-enhanced temporal adaptive reinforcement learning framework. Methodologically, it integrates variational quantum circuits (VQCs) with asynchronous advantage actor-critic (A3C) to construct an end-to-end quantum-classical hybrid policy network; additionally, it introduces a novel temporal dynamic clustering mechanism for online identification of market-state evolution and cluster-level policy transfer. Empirically evaluated on S&P 500 constituents, the model achieves 17.09% cumulative return—outperforming the benchmark by 7.09%—while demonstrating superior robustness to high-dimensional noise and enhanced exploration efficiency. Key contributions include: (i) the first quantum-enhanced A3C architecture for finance, and (ii) the first market-time-series-driven dynamic clustering paradigm for adaptive decision-making.

0 citationsRead paper
Recent publications

Latest Papers

Group Quantization and Mellin Representations of the Heston Model

Jun 11, 2026

This work provides a geometric interpretation of the Heston stochastic volatility model, uncovering the intrinsic mechanism underlying its affine structure. By constructing an augmented local Lie groupoid formulation and introducing, for the first time, group quantization techniques into this framework, the study unifies the pricing operator in coordinate space with the Riccati equations in momentum space. Leveraging the Mellin transform together with geometric representation theory, the approach not only reproduces the classical characteristic function and Riccati solutions but also demonstrates that both arise as dual perspectives of a single geometric construction. This insight endows the Heston pricing formula with a clear and coherent geometric meaning.

0 citationsRead paper

Batched Training for QLSTM vs. QFWP: A System-Oriented Approach to EPC-Aware RMSE-DA

Dec 25, 2025

This study systematically compares quantum long short-term memory (QLSTM) and quantum forgetful wave packet (QFWP) models for daily EUR/USD exchange rate forecasting under equivalent parameter count (EPC) and adjoint-based differentiation. Method: We introduce the first EPC-aligned, numerically reproducible quantum time-series modeling benchmark, incorporating batched tensor parallelism and nonparametric statistical testing (Wilcoxon signed-rank test and Cliff’s delta). Contribution/Results: QFWP consistently outperforms QLSTM across all batch sizes in RMSE and directional accuracy (p ≤ 0.004). QLSTM achieves peak throughput at batch size 64. Forward computation accelerates by 2.2–2.4×; end-to-end training acceleration reaches up to 2×. We uncover asymmetric scalability between forward and backward passes in quantum RNNs and characterize the speed–accuracy Pareto frontier. This work establishes a reproducible benchmark and provides principled model selection guidance for one-dimensional quantum time-series modeling.

0 citationsRead paper

Q-A3C2: Quantum Reinforcement Learning with Time-Series Dynamic Clustering for Adaptive ETF Stock Selection

Dec 25, 2025

To address overfitting from high-dimensional features and the inflexibility of static clustering in dynamic ETF stock selection under evolving market regimes, this paper proposes a quantum-enhanced temporal adaptive reinforcement learning framework. Methodologically, it integrates variational quantum circuits (VQCs) with asynchronous advantage actor-critic (A3C) to construct an end-to-end quantum-classical hybrid policy network; additionally, it introduces a novel temporal dynamic clustering mechanism for online identification of market-state evolution and cluster-level policy transfer. Empirically evaluated on S&P 500 constituents, the model achieves 17.09% cumulative return—outperforming the benchmark by 7.09%—while demonstrating superior robustness to high-dimensional noise and enhanced exploration efficiency. Key contributions include: (i) the first quantum-enhanced A3C architecture for finance, and (ii) the first market-time-series-driven dynamic clustering paradigm for adaptive decision-making.

0 citationsRead paper