arXiv subjects

Yeong-Jar Chang

Publications and source records attributed to Yeong-Jar Chang.

7 recordsLinked to original sources

arXiv2026

A Digital Spreading Framework for Quantum Expectation Computation Without Rotation Gates or Arithmetic Circuits

In the pursuit of quantum advantage for financial engineering, researchers face a critical dilemma: analog rotation gates suffer from inherent 'sine-to-square' biases and error magnification, while digital arithmetic circuits (e.g., WeightedAdder) incur prohibitive quadratic complexity that exceeds NISQ capabilities. This study introduces Digital Spreading (DS), a fully digital quantum computing framework designed to resolve this trade-off. DS overcomes these limitations by utilizing a pruned Cuccaro ripple-carry architecture that avoids costly multiplication and eliminates rotation gates entirely. The proposed circuit employs integer comparison operations on superposed quantum states, mapping multi-qubit outcomes onto the probability of a single target qubit. Experiments based on a random walk model for option pricing demonstrate that DS achieves floating-point precision with a relative error as low as 0.0001%, outperforming JP Morgan's rotation-based method (1.43%), as well as ITRI's analog calibration (1.43%) and digital calibration approaches (19.14%). Overall, DS provides a compact, robust, and accurate framework for quantum weighted-average computation.

quant-ph↗

arXiv2026

Integrating AI and Quantum-Inspired Techniques for Efficient Enzyme Fermentation Optimization

This paper introduces a new method that combines Artificial Intelligence (AI) and quantum-inspired techniques to improve the efficiency of multi-variable optimization experiments. By using advanced software simulations, this approach significantly reduces the time and cost compared to traditional physical experiments. The research focuses on enzyme fermentation, demonstrating that this method can achieve better results with fewer experiments. The findings highlight the potential of this approach to more effectively identify optimal formulations, leading to advancements in enzyme fermentation and other fields that require complex optimization. Initially, the Active Ingredients (AIN) could not be improved even after 600 experiments. However, by adopting the method outlined in this paper, we were able to identify a better formula in just 405 experiments. This resulted in an increase of AIN from 8481 to 10068, representing an improvement of 18.7%.

quant-ph↗

arXiv2025

Novel Quantum Circuit Designs of Random Injection and Payoff Computation for Financial Risk Assessment

Quantum entanglement enables exponential computational states, while superposition provides inherent parallelism. Consequently, quantum circuits are theoretically capable of supporting large scale parallel computation. However, applying them to financial analysis particularly in the areas of random number generation and payoff computation remains a significant challenge. Experts generally believe that quantum computing relies on matrix operations, which are deterministic in nature without randomness. This inherent determinism makes it particularly challenging to design quantum circuits that require random number injection. JP Morgan[1] introduced the piecewise linear (PWL) approach for modeling payoff computations but did not disclose a quantum circuit capable of identifying values exceeding the strike price, suggesting a possible reliance on classical pre processing for interval classification. This paper presents an integrated quantum circuit with two key components: one for random number injection, applicable to risk assessment, and the other for direct payoff computation, relevant to financial pricing. These components are compatible with a scalable framework that leverages large scale parallelism and Quantum Amplitude Estimation (QAE) to achieve quadratic speedup. The circuit was implemented on IBM Qiskit and evaluated using 8 parallel threads and 1600 measurement shots. Results confirmed both the presence of randomness and the correctness of payoff computation. While the current implementation uses 8 threads, the design scales to 2 to the power of n threads, for arbitrarily large n, offering a potential path toward demonstrating quantum supremacy.

quant-ph↗

arXiv2025

Mixed-Signal Quantum Circuit Design for Option Pricing Using Design Compiler

Prior studies have largely focused on quantum algorithms, often reducing parallel computing designs to abstract models or overly simplified circuits. This has contributed to the misconception that most applications are feasible only through VLSI circuits and cannot be implemented using quantum circuits. To challenge this view, we present a mixed-signal quantum circuit framework incorporating three novel methods that reduce circuit complexity and improve noise tolerance. In a 12 qubit case study comparing our design with JP Morgan's option pricing circuit, we reduced the gate count from 4095 to 392, depth from 2048 to 6, and error rate from 25.86\% to 1.64\%. Our design combines analog simplicity with digital flexibility and synthesizability, demonstrating that quantum circuits can effectively leverage classical VLSI techniques, such as those enabled by Synopsys Design Compiler to address current quantum design limitations.

quant-ph↗

arXiv2025

Quantum-Chiplet: A Novel Python-Based Efficient and Scalable Design Methodology for Quantum Circuit Verification and Implementation

Analysis and verification of quantum circuits are highly challenging, given the exponential dependence of the number of states on the number of qubits. For analytical derivation, we propose a new quantum polynomial representation (QPR) to facilitate the analysis of massively parallel quantum computation and detect subtle errors. For the verification of quantum circuits, we introduce Quantum-Chiplet, a hierarchical quantum behavior modeling methodology that facilitates rapid integration and simulation. Each chiplet is systematically transformed into quantum gates. For circuits involving n qubits and k quantum gates, the design complexity is reduced from "greater than O(2^n)" to O(k). This approach provides an open-source solution, enabling a highly customized solution for quantum circuit simulation within the native Python environment, thereby reducing reliance on traditional simulation packages. A quantum amplitude estimation example demonstrates that this method significantly improves the design process, with more than 10x speed-up compared to IBM Qiskit at 14 qubits.

quant-ph↗

arXiv2024

Quantum-Train: Rethinking Hybrid Quantum-Classical Machine Learning in the Model Compression Perspective

We introduces the Quantum-Train(QT) framework, a novel approach that integrates quantum computing with classical machine learning algorithms to address significant challenges in data encoding, model compression, and inference hardware requirements. Even with a slight decrease in accuracy, QT achieves remarkable results by employing a quantum neural network alongside a classical mapping model, which significantly reduces the parameter count from $M$ to $O(\text{polylog} (M))$ during training. Our experiments demonstrate QT's effectiveness in classification tasks, offering insights into its potential to revolutionize machine learning by leveraging quantum computational advantages. This approach not only improves model efficiency but also reduces generalization errors, showcasing QT's potential across various machine learning applications.

quant-ph↗

arXiv2024

Training Classical Neural Networks by Quantum Machine Learning

In recent years, advanced deep neural networks have required a large number of parameters for training. Therefore, finding a method to reduce the number of parameters has become crucial for achieving efficient training. This work proposes a training scheme for classical neural networks (NNs) that utilizes the exponentially large Hilbert space of a quantum system. By mapping a classical NN with $M$ parameters to a quantum neural network (QNN) with $O(\text{polylog} (M))$ rotational gate angles, we can significantly reduce the number of parameters. These gate angles can be updated to train the classical NN. Unlike existing quantum machine learning (QML) methods, the results obtained from quantum computers using our approach can be directly used on classical computers. Numerical results on the MNIST and Iris datasets are presented to demonstrate the effectiveness of our approach. Additionally, we investigate the effects of deeper QNNs and the number of measurement shots for the QNN, followed by the theoretical perspective of the proposed method. This work opens a new branch of QML and offers a practical tool that can greatly enhance the influence of QML, as the trained QML results can benefit classical computing in our daily lives.

quant-ph↗